Skip to main content

Full text of "The Text of New Testament 4th Edit"

See other formats

The Text of ~tie 
Nj w Testament 

Its Transmission, Corruption, 
and Restoration 


Fourth Edition 


The Text of the New Testament 

The Text of the 
New Testament 

Its Transmission, Corruption, 
and Restoration 

Fourth Edition 

Princeton Theological Seminary 


University of North Carolina at Chapel Hill 

New York ♦ Oxford 



Oxford University Press, Inc., publishes works that further 

Oxford University's objective of excellence in research, scholarship, 

and education. 

Oxford New York 

Auckland Cape Town Dar es Salaam Hong Kong Karachi 
Kuala Lumpur Madrid Melbourne Mexico City Nairobi 
New Delhi Shanghai Taipei Toronto 

With offices in 

Argentina Austria Brazil Chile Czech Republic France Greece 
Guatemala Hungary Italy Japan Poland Portugal Singapore 
South Korea Switzerland Thailand Turkey Ukraine Vietnam 

Copyright © 1964, 1968, 1992, 2005 by Oxford University Press, Inc. 

Published by Oxford University Press, Inc. 

198 Madison Avenue, New York, New York 10016 

Oxford is a registered trademark of Oxford University Press 

All rights reserved. No part of this publication may be reproduced, 
stored in a retrieval system, or transmitted, in any form or by any means, 
electronic, mechanical, photocopying, recording, or otherwise, 
without the prior permission of Oxford University Press. 

library of Congress Cataloglng-in-Publication Data 

Metzger, Bruce Manning. 
The text of the New Testament : its transmission, corruption, and restoration / by 
Bruce M. Metzger, Bart D. Ehrman. — 4th ed. 

p. cm. 
Includes bibliographical references and index. 

ISBN-13: 978-019-516667-5 (alk. paper)— ISBN-13: 978-019-516122-9 (pbk. : alk.paper) 
ISBN 0-19-516667-1 (alk. paper)— ISBN 0-19-516122-X (pbk. : alk. paper) 
1. Bible. N.T.— Criticism, Textual. 2. Bible. N.T. — History. I. Ehrman, Bart D. II. Title. 

BS2325.M4 2005 
225.4'86— dc22 


Printing number: 987654321 

Printed in the United States of America 
on acid-free paper 


Figures xi 

Preface xiii 

Preface to the First Edition xv 

Part I. The Materials for the Textual Criticism 
of the New Testament 

Chapter 1. The Making of Ancient Books 3 

I. The Materials of Ancient Books 4 

1. Papyrus 4 

2. Parchment 8 

3. Ink Making 10 

II. The Forms of Ancient Books 1 1 

III. Ancient Scribes and Their Handiwork 16 
Added Note on Colophons 31 

IV. "Helps for Readers" in New Testament Manuscripts 33 

1. Chapter Divisions (K£<j>d\aux) 34 

2. Titles of Chapters (tItAol) 36 

3. Eusebian Canons 38 

4. Hypotheses, Bioi, Euthalian Apparatus 39 

5. Superscriptions and Subscriptions 40 

6. Punctuation 41 

7. Glosses, Scholia, Commentaries, Catenae, Onomastica 41 

8. Artistic Adornment 43 

vi Contents 

9. Cola and Commata 45 

10. Neumes 46 

11. Lectionary Equipment 46 

V. Statistics of Greek Manuscripts of the New Testament 47 

Chapter 2. Important Witnesses to the Text of 
the New Testament 52 

I. Greek Manuscripts of the New Testament 52 

1 . Important Greek Papyri of the New Testament 53 

2. Important Greek Majuscule Manuscripts of 
the New Testament 62 

3. Important Greek Minuscule Manuscripts of 
the New Testament 86 

4. Other Noteworthy Manuscripts 92 

II. Ancient Versions of the New Testament 94 

1. The Syriac Versions 96 

2. The Latin Versions 100 

3. The Coptic Versions 110 

4. The Gothic Version 115 

5. The Armenian Version 117 

6. The Georgian Version 118 

7. The Ethiopic Version 119 

8. The Old Slavonic Version 121 

9. Other Ancient Versions 122 

III. Patristic Quotations from the New Testament 126 

Part II. The History of New Testament Textual 
Criticism as Reflected in Printed Editions of 
the Greek Testament 

Chapter 3. The Precritical Period: The Origin and Dominance 
of the Textus Receptus 137 

I. From Ximenes and Erasmus to the Elzevirs 137 

II. The Collection of Variant Readings 152 

Contents vii 

Chapter 4. The Modern Critical Period: From Griesbach 
to the Present 165 

I. The Beginnings of Scientific Textual Criticism of 

the New Testament 165 
II. The Overthrow of the Textus Receptus 170 

Part HI. The Application of Textual Criticism 
to the Text of the New Testament 

Chapter 5. The Origins of Textual Criticism 
as a Scholarly Discipline 197 

Chapter 6. Modern Methods of Textual Criticism 205 

I. The Classical Method of Textual Criticism 205 
II. Reactions Against Classical Textual Criticism 210 

1. Joseph Bedier 210 

2. Albert C. Clark 212 

III. Local Texts and Ancient Editions: Burnett Hillman 
Streeter 214 

IV. Alternative Methods of Textual Criticism 218 

1. The Majority Text 218 

2. Thoroughgoing Eclecticism 222 

V. Conjectural Emendation 226 

VI. Methods of Determining Family Relationships Among 
Manuscripts 231 

1 . The Claremont Profile Method 236 

2. The Alands' Use of Teststellen 237 

3. The Comprehensive Profile Method 238 

VII. The Use of Computers in New Testament 
Textual Criticism 240 

1. The Collection, Recording, and Storage of Data 242 

2. The Presentation of the Data 244 
3- Statistical Analyses 245 

4. Hypertext Possibilities 245 
VIII. Significant Ongoing Projects 246 

1 . The Institut fur neutestamentliche Textforschung 247 

2. The International Greek New Testament Project 248 

viii Contents 

Chapter 7. The Causes of Error in the Transmission 
of the Text of the New Testament 250 

I. Unintentional Changes 251 

1. Errors Arising from Faulty Eyesight 251 

2. Errors Arising from Faulty Hearing 254 

3. Errors of the Mind 257 

4. Errors of Judgment 258 
II. Intentional Changes 259 

1. Changes Involving Spelling and Grammar 26l 

2. Harmonistic Corruptions 262 

3- Addition of Natural Complements and Similar Adjuncts 263 

4. Clearing Up Historical and Geographical Difficulties 264 

5. Conflation of Readings 265 

6. Alterations Made Because of Doctrinal Considerations 265 

7. Addition of Miscellaneous Details 268 

Chapter 8. History of the Transmission of the Text 
of the New Testament 272 

I. Complications in Establishing the Original Text 272 

II. Dissemination of Early Christian Literature 274 

III. The Rise and Development of the New Testament 
Text Types 276 

1. The Western Text 276 

2. The Alexandrian Text 277 

3. The Byzantine Text 279 

IV. The Use of Textual Data for the Social History of Early 
Christianity 280 

1. Doctrinal Disputes of Early Christianity 282 

2. Jewish-Christian Relations 287 

3. The Oppression of Women in Early Christianity 288 

4. Christian Apologia 290 

5. Christian Asceticism 294 

6. The Use of Magic and Fortune-Telling in Early 
Christianity 295 

Contents ix 

Chapter 9- The Practice of New Testament 
Textual Criticism 300 

I. Basic Criteria for the Evaluation of Variant Readings 300 

II. The Process of Evaluating Variant Readings 305 

1. External Evidence 305 

2. Internal Evidence 313 

III. The Textual Analysis of Selected Passages 316 

Bibliography 245 

General Index 249 

Index of New Testament Passages 2 &3 



1 Harvesting papyrus in ancient Egypt 6 

2 Greek majuscule script 19 

3 Greek minuscule script 20 

4 Limestone statuette of Egyptian scribe 28 

5 Greek Gospel Lectionary 562 35 

6 Codex Basiliensis 37 

7 Chester Beatty Biblical Papyrus II (j) 46 ) 42 

8 Bodmer Papyrus XIV (p 75 ) 60 

9 Codex Sinaiticus GO 66 

10 Codex Bezae (D) 72 

11 Codex Laudianus (E a ) 75 

12 Codex Mosquensis (K ap ) 78 

13 Codex Sangallensis (A) 82 

14 Codex Rossanensis (S), Christ and Barabbas before Pilate 85 

15 Parchment leaf of Book of Revelation (0169) 93 

16 Curetonian Syriac MS. 97 

17 Codex Gigas 104 

18 Codex Amiatinus, Ezra the scribe 107 

19 Chester Beatty Coptic (Sahidic) MS. B 111 

20 Gothic Codex Argenteus 116 

21 Armenian Gospel MS. 229, Etchmiadzin 118 

22 Complutensian Polyglot Bible 141 

23 Greek Gospel MS. 2, University Library, Basle 144 

24 Editorial Committee, United Bible Societies 193 

25 Greek Gospel MS. 274, Bibliotheque Nationale, Paris 324 


During the 40 years that have elapsed since the initial publication of 
this book in 1964, not only have many additional Greek and ver- 
sional manuscripts of the New Testament come to light but also a 
quite considerable amount of text-critical research has been pub- 
lished in Europe and North America. Furthermore, fresh interest in 
the subject matter as well as the methodology has opened up, such 
as the contribution of computer technology in the collection and 
evaluation of manuscript evidence and the bearing of social and ide- 
ological influences upon the work of scribes. 

Room for the addition of important bibliographical items, along 
with expanded information on the making and copying of books in 
antiquity and on the history of the transmission of the text of the 
New Testament, has been gained by the elimination of materials that 
seemed to be of peripheral interest to present-day readers. English 
translation of passages of the Greek text, given previously in the Re- 
vised Standard Version of 1952, have been adjusted to the wording 
of the New Revised Standard Version of 1990. The individual illus- 
trations, formerly gathered together in one section, now are distrib- 
uted to appropriate pages throughout the volume. 

Among the people who have assisted us in our work, special 
thanks are due to Carl Cosaert, for helping to prepare the manuscript 
for publication. 

Bruce M. Metzger 
Princeton, New Jersey 

Bart D. Ehrman 
Chapel Hill, North Carolina 

November 2004 

Preface to the First Edition 

The necessity of applying textual criticism to the books of the New 
Testament arises from two circumstances: (a) none of the original 
documents is extant, and (b) the existing copies differ from one an- 
other. The textual critic seeks to ascertain from the divergent copies 
which form of the text should be regarded as most nearly conform- 
ing to the original. In some cases the evidence will be found to be 
so evenly divided that it is extremely difficult to decide between two 
variant readings. In other instances, however, the critic can arrive at 
a decision based on more or less compelling reasons for preferring 
one reading and rejecting another. 

It is the purpose of this book to supply the student with infor- 
mation concerning both the science and the art of textual criticism as 
applied to the New Testament. The science of textual criticism deals 
with (a) the making and transmission of ancient manuscripts, (b) the 
description of the most important witnesses to the New Testament 
text, and (c) the history of the textual criticism of the New Testament 
as reflected in the succession of printed editions of the Greek Testa- 
ment. The art of textual criticism refers to the application of reasoned 
considerations in choosing among variant readings. The results of 
the practice of textual criticism have differed from one generation to 
another, partly because the balance in the quantity and the quality of 
witnesses available has gradually altered owing to the acquisition of 
additional manuscripts, and partly because theories and procedures 
of evaluating textual evidence have varied over the years. In Part 
Three of the volume the author has attempted to provide a succinct 
account of the several schools of textual methodology, indicating at 
the same time what in his view is the most satisfactory critical pro- 
cedure for beginners to follow. 


xvi Preface to the First Edition 

The author gratefully acknowledges the courtesy of the follow- 
ing for granting permission to reproduce specimen folios and dia- 
grams of manuscripts: Bibliotheque Bodmer, the Bodleian Library, 
the British Museum, the Cambridge University Press, Dr. W. H. P. 
Hatch, Macmillan & Co., Ltd., and the Speer Library of Princeton 
Theological Seminary. 

Finally, I should like to express my gratitude to the Delegates of 
the Oxford University Press for their acceptance of my book for pub- 
lication. I am also indebted to the readers of the Press for their cus- 
tomary care and painstaking vigilance in the reading of the proofs. 

Bruce M. Metzger 
Princeton, New Jersey 

August 1963 

The Text of the New Testament 




The Materials for the Textual 
Criticism of the New Testament 


The Making of Ancient Books 

Until the invention of printing with movable type in the fifteenth cen- 
tury, the text of the New Testament — and, indeed, the text of every 
ancient record — could be transmitted only by laboriously copying it 
letter by letter and word by word. The consideration, therefore, of 
the processes involved in the making and transcribing of manuscripts 
is of the utmost importance to the historian of ancient culture in gen- 
eral and to the student of the New Testament in particular. The fol- 
lowing sections deal with those aspects of Greek paleography 1 that 
bear upon the textual criticism of the New Testament. 

1. Standard works on Greek paleography include Viktor Gardthausen, 
Griechische Palaeographie, 2 vols., 2te Aufl. (Leipzig, 1911-13); E. M. 
Thompson, An Introduction to Greek and Latin Palaeography (Oxford, 
1912, repr. 1975); Alfonse Dain, Les manuscrits (Paris, 1949; 3rd ed. 1975); 
N. G. Wilson, Mediaeval Greek Bookhands (Oxford, 1972-3); Ruth Barbour, 
Greek Literary Hands A. D. 400-1600 (Oxford, 1981); K. and S. Lake, 
Dated Greek Minuscule Manuscripts to the Year 1200, 10 vols. {Monu- 
menta Palaeographica Vetera, First Series, Parts I-X; Boston, 1934-9); 
Alexander Turyn, Codices Graeci Vaticani Saeculis xiii et xiv scripti anno- 
rumque notis instructi (Vatican City, 1964); idem, Dated Greek Manuscripts 
of the Thirteenth and Fourteenth Centuries in the Libraries of Italy, 2 vols. 
(Urbana, IL, 1972); D. Harlfinger, Specimina griechischer Kopisten und 
Schriftstilen des 15. und 16. fahrhunderts (Berlin, 1974); Herbert Hunger, 

4 The Text of toe New Testament 

I. The Materials of Ancient Books 

Clay tablets, stone, bone, wood, leather, various metals, potsherds 
(ostraca), papyrus, and parchment (vellum) were all used in antiquity 
to receive writing. Among these several materials, the student of the 
New Testament is interested chiefly in the last two, for almost all New 
Testament manuscripts are made of either papyrus or parchment. 

1 . Papyrus 2 

Papyrus is an aquatic plant that grows most successfully in the 
still shallows of marshlands (see Job 8.11, "Can papyrus grow where 
there is no marsh?"). Its broad root stretches horizontally under the 
mud, and from this rise several strong stalks, triangular in cross 
section; short brown leaves protect the base. Papyrus is by far the 
tallest of the botanical Cyperus papyrus, growing to a height of 12 or 
15 feet. At its top the stalk splits into a mass of strands (the umbel), 
and at the end of these the plant produces small brown flowers. The 
stalk of the papyrus plant has a tough green rind that contains an 
ivory white pith, which carries water and sustenance from the root 
to the flowering head. 

The papyrus plant was important in everyday life and was put to 
numerous practical uses. Surviving examples include sandals, bas- 
ketry, and rope, which employed the tougher parts of the plant. 

Repertorium der griechiscben Kopisten 800-1600 (Vienna, 1981); 
G. Cavallo and H. Muehler, Greek Book-hands of the Early Byzantine 
Periods A.D. 300-800 (University of London Institute of Classical Studies 
Bulletin, Supplement Papers 47, 1987); and E. G. Turner, Greek Manu- 
scripts of the Ancient World, 2nd ed., revised by P. J. Parsons (London, 
1987), with an excellent bibliography and a well-chosen album of plates 
with commentary. 

2, See Jaroslav Cerny, Paper and Books in Ancient Egypt, an inaugural 
lecture delivered at University College London 29 May 1947 (published for 
the college by H. K. Lewis & Co., London); A. Lucas, Ancient Egyptian 
Materials and Industries, 4th ed. (London, 1962), revised by J. R. Harris; 
Eric G. Turner, The Papyrologist at Work (Durham, NC, 1973); J- Vergotte, 
"L'etymologie du mot 'papyrus,'" Chronique d ' Egypte, lx (1977), 
pp. 393-7; E. G. Turner, Greek Papyri: An Introduction (Oxford, 1968; 
enlarged reproduction, 1980); Richard Parkinson and Stephen Quirke, with 
contributions by Ute Wartenberg and Bridget Leach, Egyptian Bookshelf 
Papyrus (Austin, TX, 1995). 

The Making of Ancient Books 5 

Furthermore, we learn from the fifth-century B.C. Greek writer 
Herodotus that portions of the papyrus plant could be eaten, partic- 
ularly when baked in a red-hot oven {History ii.92). Although the 
majority of the population would have used the plant chiefly for such 
simple purposes, its most inventive use was as a writing surface — 
perhaps the most influential achievement of the ancient Egyptians 
(c. 3000 B.C.). Despite the abundance of the papyrus plant in the 
marshes, the techniques of manufacturing the writing material were 
almost certainly specialized. This combined with the restriction of lit- 
eracy to the elite probably led to the concentration of the production 
of papyrus sheets and rolls in a few workshops at specific sites. 

The way in which writing material was made from papyrus is 
described by Pliny the Elder in his encyclopedic Natural History 
(XIII.xxiii.74-7), but his account is neither clear nor quite correct. 3 It 
may, however, be supplemented by an examination of actual spec- 
imens preserved from antiquity, as well as from painted mural re- 
liefs from ancient Egypt. These pictures, some of them in color, are 
found in the remains of the many tombs that have been excavated 
in the necropolis of Thebes, the second most populous and impor- 
tant city after Memphis. Thebes straddled the Nile, and from about 
1540 B.C. a line of mortuary temples and tombs took shape on 
the west bank. The tomb of a deceased official named Puyemre 4 
contains scenes of various kinds of industry in which he apparently 
had some interest, including the harvesting of papyrus. This relief 
shows three men standing in a skiff that has been pushed into a bed 
of reeds by a young man wielding a pole at the rear of the skiff. The 
workman closest to the papyrus pulls out the long stems by the 
roots while the oldest workman ties them into bundles (Fig. 1). 
The feathery heads (the umbel) are cut off, but the sheathing at the 
base is retained. 

3. For a translation and discussion, see Naphtali Lewis, Papyrus in 
Classical Antiquity (Oxford, 1974), pp. 34-69, to which he issued a brief 
supplement (Papyrologica Bruxellensia 23, 1989); for further considera- 
tions, see I. H. M. Hendricks, "Pliny, Historia Naturalis XIII, 74-82, and The 
Manufacture of Papyrus," Zeitschrift fur Papyrologie und Epigraphik, xxxvii 
(1980), pp. 121-36; Harry Y. Gamble, Books and Readers in the Early 
Church (New Haven, 1995), pp. 44-5. 

4. Norman de Garis Davies, The Tomb of Puyemre at Thebes, 2 vols. 
{The Metropolitan Museum of Art; New York, 1922-3), vol. i, "The Hall of 
Memories," pp. 61-6 and Plates xviii and xix. 

The Text of the New Testament 




^^™ * 


Figure 1 Harvesting papyrus in ancient Egypt. Two parts of a colored 
painting are reproduced here in black and white from Plates xviii and xix in 
Norman de Garis Davies, The Tomb of Puyemre at Ihebes, i (New York, 
1922). It will be noted that the pole used by the man rowing the skiff links 
visually the two plates in Davies' publication. 

The Making of Ancient Books 7 

Further to the right of the same relief (which is presented here 
in two sections), a man carries a bundle of the stems on his back, 
evidently supplying another, who is seated on a low stool busily 
working on the stems and quite possibly preparing them for manu- 
facture into writing material. He is holding one end of a stem, the 
flower of which has been cut away, between the toes of his left foot 
and the other end in the fingers of his left hand, while, with his right 
hand, he is stripping the stem of its outer rind. 

There is no continuation of the representation in the tomb, but 
we know that the stem was first stripped of its outer rind and then 
cut into shorter pieces about 18 or 20 inches in length. The inner pith 
was subsequently cut into thin, narrow slices. All this preparation 
was done while the plant was still fresh and green. 

To make a sheet of papyrus, these slices were placed vertically 
side by side on a hard wooden plank or table with their edges 
slightly overlapping. On this first layer another was put horizontally, 
with the slices running at right angles to the slices of the first layer. 
By pressing and beating, the two layers became one sheet, the 
plant's natural juice gluing the layers firmly together; the sheet thus 
made was dried under pressure. Lastly, the surface was polished 
with some rounded object, possibly of stone, until it became per- 
fectly smooth. The borders were then cut in order to make them 
straight and to give the sheet a rectangular shape. The upper layer 
showed horizontal slices or, as we call them, fibers running left to 
right, while the lower layer consisted of vertical fibers. 

Though the length of papyrus stems could have permitted the 
manufacture of sheets of considerable size, the sheets usually had 
a maximum height of about 15 inches and a maximum breadth of 
about 9 inches. To create a larger writing surface than that 
provided by a single sheet, a number of sheets were pasted to- 
gether along their height, each overlapping the sheet on its left 
by about 'A to 1 inch. This pasting together (using a common 
starch paste) was done at the factory, where the number of sheets 
that formed a standard roll was twenty (see Pliny, Natural His- 
tory, XIII.xxiii.77). 

After obtaining a blank roll from the stationer, the scribe would 
cut off whatever length was deemed to be necessary for the job at 
hand, whether a letter or a literary work. If, however, a still longer 
roll was required, the scribe could, of course, paste another roll to 
the original roll. 

8 The Text of the New Testament 

2. Parchment 5 

The manufacture of parchment for writing purposes has an 
interesting history. According to Pliny the Elder {Natural History, 
xiii.21 f), it was King Eumenes of Pergamum, a city in Mysia of Asia 
Minor, who promoted the preparation and use of parchment. This 
ruler (probably Eumenes II, who ruled 197-159 B.C.) planned to 
found a library in his city that would rival the famous library of 
Alexandria. This ambition did not please Ptolemy of Egypt (probably 
Ptolemy Epiphanes, 205-182 B.C.), who thereupon put an embargo 
on the export of papyrus sections. It was this embargo that forced 
Eumenes to develop the production of vellum, which from the place 
of its origin received the Greek name jTEoyafinvij (whence our Eng- 
lish word parchment is derived). Whatever may be thought of the de- 
tails of this story — actually, leather (parchment) was used for books 
long before Eumenes — the gist of it is probably true, namely that a 
high quality of parchment was developed at Pergamum and that the 
city became famous for the manufacture and export of this kind of 
writing material, eventually giving its name to the product. 

Parchment, however, did not come into general use for book 
production until some centuries later, even though it had a marked 
advantage over papyrus in its greater durability; moreover, it was 
better suited than papyrus for writing on both sides. It was at about 
the start of the fourth century A.D. that it began to take the place of 
papyrus in the manufacture of the best books, and the works con- 
sidered worth preserving were gradually transferred from papyrus 
roll to parchment codex. It is in this century that the great parchment 
codices of the Greek Bible (the Vaticanus and the Sinaiticus; see 
pp. 62-7) were prepared; and the earliest extant parchment manu- 
scripts of pagan works date probably from the same century. But 
the use of papyrus did not cease then, and papyrus manuscripts of 
the New Testament have been found dating from the fifth, sixth, and 
seventh centuries. 

5. See Karl J. Liithi, Das Pergament, seine Geschichte, seine Anwen- 
dung (Bibliothek des Schweizerischen Gutenberg-museums, vol. vi; 
Bern, 1938); R. Reed, Ancient Skins, Parchments, and Leathers (New York, 
1972); Christopher de Hamel, Medieval Craftsmen; Scribes and Illuminators 
(Toronto, 1992); M. L. Ryder, "The Biology and History of Parchment," in 
Pergament: Geschichte, Structur, Restaurierung, Herstellung, ed. by P. Ruck 
(Sigmaringen, 1991), pp. 25-33; and Gamble, Books and Readers, pp. 45-6. 

The Making of Ancient Books 9 

Parchment or vellum (the two words are often used interchange- 
ably, but exact writers restrict the word vellum to describe a finer, su- 
perior quality of parchment) was made from the skins of cattle, sheep, 
goats, and antelopes, especially from the young of these animals. 

The preparation of parchment, according to Christopher de 
Hamel's lengthy account 6 (which is condensed here), was a slow 
and complicated process. Early manuals emphasize that the selection 
of good skins is crucial. After skins with unacceptable flaws have 
been discarded, the parchmenter has to remove the hairs. This is 
accomplished by soaking the skins in wooden or stone vats in a so- 
lution of slaked lime for 3-10 days (longer in winter, apparently), 
stirring the vats several times a day with a wooden pole. 

One by one the wet, slippery skins are scooped out and draped, 
hair side out, over a great curved upright shield of wood, called a 
beam. The parchmenter stands behind the beam, leaning forward 
over the top, and scrapes away the hair with a long, curved knife 
with a wooden handle at each end. The skin is still very wet and 
drippy with the lime solution. It is then flipped over on the wooden 
frame so that the original inner side is outermost. The parchmenter 
once again leans over the board and with the curved knife pares 
away the residue of clinging flesh. The dehaired and tidied-up pelt 
is then rinsed for two days in fresh water to clear it of the lime. This 
completes the first and smelliest stage of parchment making. 

In the second phase of the process, the skin is actually made 
into parchment. This involves drying the skin while it is stretched 
taut on a wooden frame. The skin cannot be nailed to the frame be- 
cause as it dries it shrinks and the edges would tear away. So instead, 
the parchmenter suspends the skin by short strings attached to ad- 
justable pegs in the frame. Every few centimeters around the edge of 
the skin the parchment maker pushes little pebbles or smooth stones 
into the soft border, folding them in so as to form knobs, each of 
which is then looped around and secured with cord. The other end 
of the cord is then anchored into the slot of a revolving peg in 
the frame. One by one these knobs and strings are lashed around the 
edge until the whole skin resembles a vertical trampoline, and the 
pegs are turned to pull the skin taut. 

The skin is now tight and rubbery but still wet. The parchment 
maker begins scraping vigorously at the skin using a curved knife with 

6. de Hamel, op. cit., pp. 8-16. 

10 The Text of the New Testament 

a central handle. The crescent-shaped knife, called a lunellum, occurs 
in medieval pictures of parchment makers as their most recognizable 
tool, used to give both surfaces a thorough scraping. As the work pro- 
gresses, the parchmenter constantly lightens the pegs and taps them 
with a hammer to keep them fixed. Then, the taut skin is allowed to 
dry on the frame, perhaps helped by exposing it to the sun. 

When it is all dry, the scraping begins again. The amount of 
scraping will depend on the fineness of the parchment being made. 
In the early monastic period of manuscript production, parchment 
was often quite thick, but by the thirteenth century it was being 
planed away to an almost tissue thickness. 

Now the pegs can be undone. The dry, thin parchment can be 
rolled up or taken to be sold. Probably, when medieval scribes or 
booksellers bought vellum from a percamenarius, it was like this, not 
yet buffed up and rubbed with chalk in preparation for actual writing. 

Vellum or parchment continued to be generally used until the 
late Middle Ages. At that time, paper, which was made of cotton, 
hemp, or flax, having been introduced into Europe from China by 
Arabian traders, became popular and supplanted other writing mate- 
rials. Exactly how quickly the use of paper spread is uncertain, but it 
is clear that by the eleventh century Byzantine scribes were using it 
frequently and by the middle of the fourteenth century parchment 
had been largely replaced by paper, much of it produced in western 
Europe and datable within fairly narrow limits thanks to the evidence 
in the watermarks. 

3. Ink Making 

The most common ink used in writing on papyrus was a carbon- 
based mixture made of charcoal or lampblack and water, with some 
ground-up gum arabic — not so much to supply adhesive as to make 
the ink thicker. Since carbon inks do not stick well to parchment, 
another kind was developed, using oak galls and ferrous sulfate, 
known also as "copperas." An oak gall is a curious ball-like tumor, 
about the size of a small marble, that grows mainly on the leaves or 
twigs of oak trees. It is formed when the gall wasp lays its egg in the 
growing bud of the tree and a soft, pale green, apple-like sphere be- 
gins to form around the larva. When the larva inside is fully developed 
into an insect, it bores a hole out of its vegetable cocoon and flies 
away; the hard nut that remains is rich in tannic and gallic acids. Such 
nuts then are roughly crushed and infused for some days in rainwater. 

The Making of Ancient Books 1 1 

The ferrous sulfate is then added to the oak-gall potion, and the re- 
sulting solution slowly turns from pale brown into black ink; some 
gum arabic is added to make the ink thicker. Iron-gall ink darkens 
even further when exposed to air on the pages of a manuscript. 

Deluxe editions of manuscripts, according to St. Jerome, who 
did not approve of such extravagance, 7 were made of vellum dyed a 
deep purple 8 and written with gold and silver inks. Ordinary editions 
were written with black or brown ink and had decorative headings 
and initial letters colored with blue or yellow or (most often) red 
ink — whence the word rubric, from ruber, the Latin for "red." 

II. The Forms of Ancient Books 

In the Greco-Roman world, literary works were customarily 
published in the format of a scroll, made of papyrus or parchment. 
The papyrus scroll was made by gluing together, side by side, sepa- 
rate sheets of papyrus and then winding the long strip around a 
roller, thus producing a volume (a word derived from the Latin 
volumen, "something rolled up"). The length of a papyrus roll was 
limited by considerations of convenience in handling it; the normal 

7. In his famous letter to Eustochium, Jerome inveighs against 
anomalous extravagance: "Parchments are dyed purple, gold is melted 
into lettering, manuscripts are decked with jewels, while Christ lies at the 
door naked and dying" (Epist. 22.32; see also Jerome's preface to the 
Book of Job and Evaristo Arns, La Technique du livre d'apres Saint 
Jerome [Paris, 19531). Writing to a correspondent named Laeta, who has 
asked how she ought to rear her young daughter, he advises, "Let her 
treasures be not gems or silks, but manuscripts of the holy Scriptures; 
and in these let her think less of gilding and Babylonian parchment and 
arabesque patterns, than of correctness and accurate punctuation" 
(Epist. 107.12). 

8. Concerning purple manuscripts, see Courtney M. Booker, "The 
'Codex Purpureas' and Its Role as 'imago regis' in Late Antiquity," in 
Studies in Latin Literature and Roman History, viii, ed. by Carl Deroux 
{Collection Latomus, 239; Brussels, 1997), pp. 441-77; Italo Furlan, 
"Introduzione a codici purpurei," in Laporpora: realta eimmaginario di 
un colore simbolico, ed. by Oddone Longo (Venice, 1998), pp. 317-37. For 
a discussion of the special skills required to produce a codex aureus 
purpureus, see E. A. Lowe in Studies in Art and Literature for Belle da 
Costa Greene, ed. by Dorothy Miner (Princeton, 1954), pp. 266-8. 

12 The Text of the New Testament 

Greek literary roll seldom exceeded 35 feet in length. 9 Ancient 
authors therefore would divide a long literary work into several 
"books," each of which could be accommodated in one roll. The two 
longest books in the New Testament, the Gospel of Luke and the 
Book of Acts, would each have filled an ordinary papyrus roll of 31 
or 32 feet in length. Doubtless, this is one of the reasons why Luke 
and Acts were issued in two volumes instead of one. 

On the roll thus formed the writing was arranged in a series of 
columns, each about 2 or 3 inches wide. The height of the columns, 
which ran parallel to the stick on which the roll was wound, varied, 
of course, with the height of the original papyrus sheets. Sometimes, 
but not very often, the roll was written on both sides (see Rev. 5.1); 
this was called an opisthograph. 

The roll was relatively inconvenient to use. Two hands were 
needed to hold it open, to wind it so that its narrow columns of 
writing should always be within the reader's angle of vision, and 
after use to rewind it. Athenian vase paintings 10 show readers get- 
ting into difficulties with a twisted roll, and the aged Verginius Rufus 
broke his hip while trying to collect up one he had dropped (Pliny, 
Epistle 2.1.5). 

Moreover, the Christian community soon discovered how labori- 
ous it was to try to find specific passages in their sacred books when 
they were written in roll form. Early in the second century (or per- 
haps even at the close of the first century), the codex, or leaf form of 
book, came into use in the Church. A codex was made by folding 
one or more sheets of papyrus in the middle and sewing them to- 
gether. Christians found that this form had a number of advantages 
over the roll: (1) it permitted all four Gospels or all the Epistles of 
Paul to be bound into one book, a format that was impossible so 

9. Callimachus, the learned cataloguer of books in the great library at 
Alexandria, was accustomed to say, "A big book is a big nuisance" {fieya 
fiifiAiov fxiya xaxov, fr. 465). Ceremonial copies of the Egyptian Book of 
the Dead have been found that measure more than 100 feet in length; 
these were not meant to be read but to be buried in the tomb of a rich 

10. J. D. Beazley, American Journal of Archaeology, Hi (1948), 
pp. 336-40; H. R. Immerwahr, American Journal of Archaeology, lxix 
(1965), pp. 152 ff.; idem, "Book Rolls on Attic Vases," in Studies in Honor 
ofB. I. IJllman, ed. by H. R. Immerwahr (Rome, 1964), pp. 17^8. 

The Making of Ancient Books 13 

long as the roll was used; (2) it facilitated the consultation of proof 
texts; and (3) it was better adapted to receiving writing on both sides 
of the page, thus keeping the cost of production down. The sugges- 
tion has been made that it was Gentile Christians who early adopted 
the codex form for their Scriptures instead of the roll form as part 
of a deliberate attempt to differentiate the usage of the Church 
from that of the synagogue, which was accustomed to transmit the 
Old Testament on scrolls. 11 Other scholars have maintained that 
the codex became popular among Christians because of the influ- 
ence exerted on the entire tradition by some particularly authorita- 
tive sacred text or texts — either the Gospel of Mark, the four Gospels 
together, or the episdes of Paul. 12 

In any event, it is clear that it was the Christians who early 
adopted and popularized the format of the codex in preference to 
the scroll for their sacred books. Whereas among surviving Greek 
manuscripts of pagan texts, whether literary or scientific writings, 
only 14 of 871 items dated to the second century A.D. are in the form 
of a codex, all the surviving Christian biblical papyri of the same 
period are in codex format (11 in number). Of the approximately 
172 biblical manuscripts or fragments of manuscripts written before 
A.D. 400 or not long thereafter, it appears that 158 come from 
codices and only 14 from rolls. During the same period, the codex 
was preferred also for nonbiblical Christian literature; of 118 such 

1 1 . See Peter Katz, "The Early Christians' Use of Codices Instead of 
Rolls," Journal of Theological Studies, xliv (1945), pp. 63-5. For a different 
view, see Saul Lieberman, "Jewish and Christian Codices," in Hellenism in 
Jewish Palestine (New York, 1950), pp. 203 ff., who suggests that "the first 
Jewish Christians, such as Matthew and Mark, would follow the accepted 
Jewish practice and put down their VTtOftvtjfiaza in codices," for codices 
or notebooks were regarded as "the most suitable way of indicating that 
they were writing the Oral Law for private, or unofficial use, and not for 
publication" (p. 205). See also C. H. Roberts, "P. Yale 1 and the Karly 
Christian Book," in Essays in Honor of C. Bradford Welles {American 
Studies in Papyrology, vol. 1; New Haven, 1966), pp. 25-9. 

12. Thus, C. II. Roberts, "The Codex," Proceedings of the British 
Academy, xl (1954), pp. 169-204; T C. Skeat, "The Origin of the Christian 
Codex," Zeitschrift fur Papyrologie und Epigraphik, cii (1994), pp. 263-8 
(all four Gospels); Gamble, Books and Readers, pp. 49-66 (an early 
collection of Paul's letters). 

14 The Text of the New Testament 

texts, 83 are from codices and the remaining 35 are rolls. 1 ' There 
seems, therefore, to have been a remarkable uniformity in the prac- 
tices of Christian scribes from the earliest times in preferring the 
codex format. 

The economic advantage of making a book in the format of 
codex instead of scroll has often been noticed, but it is only recently 
that the amount of saving has been rather precisely calculated. Ac- 
cording to Skeat, the saving in the cost of papyrus when one used 
the format of codex instead of roll was about 44%. When a book was 
produced by a commercial scribe, who would, of course, charge the 
same amount for writing in either case, the combined cost of writing 
and papyrus would have presented a saving of about 26% by chang- 
ing from roll to codex. 14 

The advantages of parchment over papyrus for the making of 
books seem obvious to us today. It was a much tougher and more 
lasting material than papyrus. Moreover, parchment leaves could 
receive writing without difficulty on both sides, whereas the vertical 
direction of the fibers on the "reverse" side of a sheet of papyrus 
made that side somewhat less satisfactory than the other as a writing 
surface. On the other hand, parchment also had its defects. For ex- 
ample, the edges of parchment leaves are apt to become puckered 
and uneven. Furthermore, according to Galen, the famous Greek 
physician of the second century A.D., parchment, which is shiny, 

13. For these statistics, see C. H. Roberts and T. C. Skeat, The Birth of 
the Codex (London, 1983), pp. 37-43, being an enlarged version of "The 
Codex" by Roberts, op. cit. They suggest that the adoption of the papyrus 
codex by Christians probably took place at Antioch not later than c. A.D. 
100 (pp. 58-61). For the question of whether the parchment codex did or 
did not antedate the papyrus codex, reference may be made to the discus- 
sion in E. C. Turner, The Typology of the Early Codex (Philadelphia, 1977), 
pp. 35-42; the two may have developed in parallel. See also Gamble, 
Books and Readers, pp. 49-66, along with several contributions in the 
colloquium Les debats du codex, ed. by Alain Blanchard (Turnhout, 1989). 

14. T. C. Skeat, "The Length of the Standard Papyrus Roll and the Cost- 
Advantage of the Codex," Zeitschrift fur Papyrologie und Epigraphik, xlv 
(1982), pp. 169-75. On the cultivation and harvesting of the papyrus plant, 
as well as all aspects of its commercial uses, see Lewis, Papyrus in Classi- 
cal Antiquity. According to William V. Harris, the cost of papyrus as writing 
material was quite prohibitive for the poorer classes in the Greco-Roman 
world {Ancient Literacy [Cambridge, MA, 19891, pp. 194 f., 239). 

The Making of Ancient Books 15 

strains the eyes more than does papyrus, which does not reflect so 
much light. 

Eusebius, the noted Christian scholar of Caesarea in Palestine, in- 
cluded in his Life of Constantine information concerning an imperial 
request for 50 parchment manuscripts. About A.D. 331, when Con- 
stantine wished to secure copies of the Scriptures for the new churches 
that he proposed to build in Constantinople, he wrote to Eusebius re- 
questing him to arrange without delay for the production of "fifty 
copies of the sacred Scriptures ... to be written on fine parchment in 
a legible manner, and in a convenient portable form, by professional 
scribes (xaXXiygd<pot) thoroughly accomplished in their art." 15 These 
orders, Eusebius continues, "were followed by the immediate execu- 
tion of the work itself, which we sent him in magnificent and elabo- 
rately bound volumes of threefold and fourfold forms." 16 

The suggestion has been made by several scholars that the two 
oldest parchment manuscripts of the Bible that are in existence today, 
namely Codex Vaticanus and Codex Sinaiticus (see Chapter 2 for 
descriptions of these manuscripts), may have been among those 
ordered by Constantine. It has been pointed out that Eusebius' curi- 
ous expression "volumes of threefold and fourfold forms" agrees 
with the circumstance that these two codices have, respectively, three 

15. Eusebius, Life of Constantine, iv. 36. 

16. The Greek text of the concluding clause (iv jtoXvxeXmg 
r\ay.nfiivoic, zevyEaiv rgiooa xal rergaoaa diane/xipavrcuv fi^iciv) is difficult 
to interpret, and the words rgiooa xal rergaoad have been taken in widely 
different senses. Thus, it has been suggested that the words refer to codices 
that were composed of quires of three or four double leaves; that they were 
polyglot Bibles in three or four languages; that they were harmonies of 
three or four Gospels; that copies were sent off to Constantine three or four 
at a time; that each Bible was in three or four parts; and that the pages had 
three or four columns of script. Each of these interpretations involves more 
or less serious difficulties; perhaps the least unsatisfactory interpretation is 
the one mentioned last. For discussions of the problems involved, see Kir- 
sopp Lake, "The Sinaitic and Vatican Manuscripts and the Copies Sent by Eu- 
sebius to Constantinople," Harvard Theological Review, xi (1918), pp. 32-5; 
J. H. Ropes, The Textof Acts (The Beginnings of Christianity, part I, vol. iii; 
London, 1926), pp. xxxvi ff.; Carl Wendel, "Der Bibel-Auftrag Kaiser 
Konstantins," Zentralblatt fiir Bibliothekswesen, lvi (1939), pp. 165-75; T. C. 
Skeat, "The Use of Dictation in Ancient Book-Production," Proceedings of 
the British Academy, xlii (1956), pp. 196 f. 

16 The Text of ran New Testament 

columns and four columns on each page. There are, however, one or 
two indications that point to Egypt as the place of origin of Codex 
Vaticanus, and the type of text found in both codices is unlike that 
used by Eusebius. The most that can be said with certainty, therefore, 
is that Codices Vaticanus and Sinaiticus are doubtless like those that 
Constantine ordered Eusebius to have copied. 

III. Ancient Scribes and Their Handiwork 

In writing on papyrus, the scribe was accustomed to utilize the 
horizontal fibers on the recto 17 side of the sheet as guidelines for the 
script. Before writing on parchment, the surface would be marked 
with a blunt-pointed instrument, drawing not only horizontal lines but 
two or more vertical lines as well, thus indicating the margins of each 
column of writing. In many manuscripts, these lines are still visible, as 
are also the pinpricks that the scribe made first as a guide for ruling 
the vellum. 18 Different schools of scribes employed various proce- 
dures of ruling, and occasionally it is possible for the modern scholar 
to identify the place of origin of a newly discovered manuscript by 
comparing its ruling pattern (as it is called) with those in manuscripts 

17. The terms recto and verso are ambiguous; when used by papyrol- 
ogists, they mean, respectively, the side on which the fibers are horizontal 
and that on which they are vertical, whereas in describing manuscripts, 
recto means right-hand page when the codex is opened and verso, the 
reverse side of that page. 

18. There is even a science of pinpricks! See E. K. Rand, "Prickings in 
a Manuscript of Orleans," Transactions and Proceedings of the American 
Philological Association, lxx (1939), pp. 327-41; L. W. Jones, "'Pin Pricks' 
at the Morgan Library," Transactions and Proceedings of the American 
Philological Association, lxx (1939), pp. 318-26; idem, "Where are the 
Prickings?" Transactions and Proceedings of the American Philological 
Association, lxxii (1944), pp. 71-86; idem, "Pricking Manuscripts: The 
Instruments and Their Significance," Speculum, a Journal of Mediaeval 
Studies, xxi (1946), pp. 389-403; idem, "Prickings as Clues to Date and 
Origin: The Eighth Century," Medievalia et humanistica, xiv (1962), 

pp. 15-22. Rand dealt earlier with various methods of ruling manuscripts 
in vogue during the Middle Ages; see his study "How Many Leaves at a 
Time?" Palaeograpbica Latina, v (1927), pp. 52-78. 

The Making of Ancient Books 17 

whose place of origin is known. 19 Since the hair side of vellum is 
darker than the flesh side, it was found that the most pleasing effect 
upon the reader was obtained only when the separate sheets were not 
indiscriminately gathered together in quires, but when the hair side of 
one page faced the hair side of the opposite page and the flesh side 
faced the flesh side, wherever the book was opened. 20 

In antiquity, two styles of script for writing Greek were in general 
use: the book-hand and the cursive. Both have always existed side by 
side; the book-hand is conservative, but the cursive can change very 
quickly, with forms that tend to invade the book-hand. The cursive, or 
"running," hand could be written rapidly and was employed for non- 
literary, everyday documents, such as letters, accounts, receipts, peti- 
tions, deeds, and the like. Contractions and abbreviations of frequently 
recurring words (such as the definite article and certain prepositions) 
were common. 21 Literary works, however, were written in the more 
formal style of book-hand, which was characterized by more deliber- 
ate and carefully executed letters, each one separate from the others — 
somewhat like writing in capital letters (not uncials, a word that has 
precise meaning in Latin writing but only a derived and imprecise one 
in Greek). 22 

Hands are also divided into majuscules and minuscules: in 
majuscules, which comprise capitals and early cursives, the letters 

19. For a list of several hundred different ruling patterns, see K. and 
S. Lake, Dated Greek Minuscule Manuscripts, index vol. ed. by Silva Lake 
(Boston, 1945). The 757 plates contain approximately 1,000 facsimiles of 
folios from 401 manuscripts in 30 different libraries. The index volume 
contains indices in 14 categories (including names of scribes). 

20. This characteristic feature of parchment codices was discovered 
toward the end of the nineteenth century by Caspar R. Gregory; see his 
article "The Quires in Greek Manuscripts," American Journal of Philology, 
vii (1886), pp. 27-32. 

21. Kathleen McNamee has gathered an extensive list in her "Abbrevi- 
ations in Greek Literary Papyri and Ostraca" (Bulletin of the American 
Society of Papyrologists, Supplement 3; Chico, CA, 1981). 

22. The word uncialis derived from the Latin uncia, meaning "a twelfth 
part" of anything. Apparently, the term came to be applied to letters that oc- 
cupied roughly one-twelfth of an ordinary line of writing. See W. H. P. Hatch, 
"The Origin and Meaning of the Term 'Uncial,'" Classical Philology, xxx 
(1935), pp. 247-54; Paul Mayvaert, "'Uncial Letters': Jerome's Meaning of the 
Terra," Journal of 'theological Studies, n.s., xxxiv (1983), pp- 185-8. 

18 The Text of the Nkw Testament 

mainly lie between two (notional) parallel lines (bilinearity). In later 
cursive, many letters with long ascenders and descenders developed, 
and these were taken over by the subsequent book-hands. Such 
hands are called minuscules, scripts in which the bodies of the letters 
lie between two inner lines but the ascenders and descenders reach 
out toward two outer lines above and below (quadrilinearity). In 
virtually all cases, only one line is actually ruled: the one on which 
the letters stand or from which, in Greek after A.D. 1000, they hang. 
Some of the most beautiful specimens of Greek handwriting are 
certain classical and biblical manuscripts dating from the third to the 
sixth centuries. In the course of time, however, the style of the book- 
hand began to deteriorate, and the letters became thick and clumsy. 
Then, about the beginning of the ninth century, a reform in hand- 
writing was initiated, and a script of smaller letters in a running hand 
was created for the production of books. 23 This modified form of 
the cursive script became popular almost at once throughout the 
Greek world, though some liturgical books continued for one or two 
centuries to be written in majuscule script. Thus, manuscripts fall 

23- The credit for initiating the reform in Greek handwriting has been 
commonly attributed to scholarly monks at the monastery of the Studium in 
Constantinople (see, e.g., G. Zereteli, "Wo ist das Tetraevangelium von Por- 
phyrius Uspenskij aus dem Jahre 835 entstanden?" Byzantinische Zeitschrift, 
ix [1900], pp. 649-53; T. W. Allen, "The Origin of the Greek Minuscule 
Hand," Journal of Hellenic Studies, xl [1920], pp. 1-12), but it has also been 
argued that the perfecting of minuscule script for book production was the 
work of humanistic scholars who were involved in the revival of culture at 
Constantinople during the second epoch of iconoclasm (see Bertrand Hem- 
merdinger, Essai sur I'histoire du texte de Thucydide [Paris, 19551, pp. 33-9). 

The earliest known minuscule Greek manuscript bearing a date is a 
copy of the four Gospels, now in the public library of St. Petersburg, with 
its all-important colophon dated 7 May 6343 (A.D. 835) by the monk 
Nicolaus, later abbot of the Studium. This manuscript has raised problems 
for scholars; paleographically, the handwriting appears to be too mature 
and fully developed to stand at the beginning of the minuscule period, yet 
no forerunners have been recognized among extant manuscripts. For a 
discussion, see Aubrey Diller, "A Companion to the Uspenski Gospels," 
Byzantinische Zeitschrift, xlix (1956), pp. 332-5. See also Barbara Aland 
and Klaus Wachtel, "The Greek Minuscule Manuscripts of the New 
Testament," in The Text of the New Testament in Contemporary Research: 
Essays on the Status Quaestionis, ed. by Ban D. Ehrman and Michael W. 
Holmes (Grand Rapids, MI, 1995), pp. 43-60. 

The Making of Ancient Books 19 

noyoYKei xeNfH d>HTi xh cxtoyn 


ti)c€2XNe:Tix<FNisi XKoycereKKi^r 


rHcMXioYAexNX RxenowrecRV.. 



M H ex 61 Nfl 2X1* rX>H KXfXJ XTOY 


AXAXxeerieceNe toiccucinxyw 
niTxcxKXNexc ftxrecucHKoMK 



XHNKXiexixoy a>ciNXKoycu>cr 


KXfpHpA^eJtH * pi n KXien |.crn» 
~^7XO*6)Xoexa>N _ ^Kocinkxmxcw* 

KxinrocKSBj*. ^VMameM>W 

Figure 2 Greek majuscule script (from Codex Sinaiiicus, fourth century; 
actual width of each column about 2'/i inches). 

Col. a, Matt. 13.5-10, aXXa de eneaev e\m xa nexgmdt] o\nov ovx ei%ev yrjv \ 
TioXXrjv xai ev8e\wg e^avexiXev bi\a xo fit] e%iv (iaOog \ ytjg tjXiov de 
ava\xiXavxog exavfia\xia6t] xai dia xo \ fit] eyeiv gi^av e\£r]gavdt] \ aXXa de 
Eiteaev e\m xaq axavOag | xai avefirjoav ai \ axavOai xai envi\^av aura | 
aXXa de eneaev e\m xtjv yt]v xr/v xa\Xt]v xai edidov \ xoqtiov ofiev t'\xaxov o 
de ei;t}xov\xa o de Xo e%o>v \ u>xa [insert axovew from the left-hand margin] 
axovexcu \ xai ngoaeXOovxeg \\. Col. b, xMatt. 13.14-16, xai avanXt]gov\rai 
avxoig rj 7i(jo\(j>r]xia tfoaiov rj \ Xeyovaa axorj | axovaexe xai ov | fit] ovvtjxe 
xai | fSXejxovreg fiXe\ipt]re xai ov fit] i\dt]xe ejtaxvv9t] \ yag t] xagdia xov \ 
Xaov xovxov xai \ xoigcuoiv avxatv \ fiagecog tjxovoav | xai rove; 0(p6aX\/iovg 
avxwv exafi\fivoav firjjxoxe \ idojow xotg o\<j>8aXfioig xai xoig \ cooiv laurwv 
between the lines] axovawaw | xai xtj xagdia ovvw\oiv xai EJiioxge\ipa>oiv 
xai iaaopts | avrovg | vfimv de fiaxagt. The Eusebian canon numerals (see 
pp. 38-9) stand between the columns IgXd _ 134\ 

V e ~ 5 


The Text of the New Testament 

lyuaifTBPaAttyTpcr TStB CD ^ uu P aB ^K**V m> 
'6y£y 4m aj-arajj-rxoy • ^t^\<rn\j-iioieott^iyna 

A"A our * oyymUaf I i-asy axrro\t r%ji * liuai 

pnVuw 6-|/«)jL4)i>aDO'f pi«-amn-ouM«Tt • ^y*r 

^a^ttUjiJuy^irrlao * 

»V U V Tcur^paAeuT ; 

• . ' . ' y 
t» «T» a» p cu i/n."? crp w 

<i^p ap o-LU/ « ^5p a f cr fu 


l»*u n La un£y asrroxi 

crp<MPf * otjD\oLxur«&f 
asra^Ury o er a*sry vcr 

Figure 3 Greek minuscule script (from Lectionary 303, twelfth century; 
actual width of each column about 3 l A inches). 

Col. a, Luke 24.31-3, xai ambg atpavrog \ eyivero djf amcbv \ xai sbtov 
Tigbg dX\krjXovg- ov%i i] xag\6ta fificbv xatofie\vrj f\v ev fi/uv. mg e\Aa?.ei r\ylv 
ev rfj6\6d> xai tog dirjvoiysv \ fifilv rag ygcupag; \ xal avaoTavxeg av\xr] xr) 
coga V7teaxgs\ipav slg [[egovoaWj/f xai \ svgov ovvt]dgoio/iE\vovg xovg 
evdsxa ||. Col. b, John 1 .35-8, Td> xaigdi exeLvcq- \ eioxtjxei 6 icudvvrjg \ xai ex 
x&v jxaQr]\x&v avrov dvo- xai \ ififlXEipag xw i[t]Oo]v jt£\giJiatovvn- Aeyer | 
ids 6 dfivog xov 6[eo\v- \ xai fjxovoav avrov \ oi dvo fiaGrjxai Xa\kovvxog- 
xai i]xo\kovQr)aav x& i[r]oo]v- | axgatpsig 6 i[rjoov]g xai d£\aodfi£vog 
avxovg \ dxoXovQovvxag. 

into two rather well-defined groups, the earlier being written in ma- 
juscule letters (see Fig. 2) and the later in minuscules (see Fig. 3). 

The advantages of using minuscule script are obvious. Minus- 
cule letters, as their name suggests, are smaller than majuscules, and 
thus the writing is more compact. Hence, when the minuscule hand 
was used, less parchment was required and therefore the book was 
more economical. Furthermore, a literary work could be produced 
that was less bulky and therefore easier to handle than a larger 
manuscript. Moreover, it was possible to write minuscule letters more 
rapidly than majuscules, and consequently books could be produced 
more quickly and more cheaply. 

The Making of Ancient Books 21 

It is easy to understand that this change in the style of script had 
a profound effect upon the transmission of Greek literature in general. 
According to Reynolds and Wilson, 24 

The transliteration of old uncial books into the new script was energet- 
ically undertaken by the scholars of the ninth century. It is largely 
owing to their activity that [classical] Greek literature can still be read, 
for the text of almost all [suchl authors depends ultimately on one or 
more books written in minuscule script at this date or shortly after, from 
which all later copies are derived; the quantity of literature that is avail- 
able to us from the papyri and the uncial manuscripts is only a small 
proportion of the whole. 

Now the possession of copies of the Scriptures (and of other lit- 
erary works) was placed within reach of persons of limited means. 
When literary works were copied almost exclusively in the majuscule 
script, such persons were obliged to get along without many books. 
Thus, the minuscule hand was an important factor in the dissemina- 
tion of culture in general and of the Scriptures in particular. The mi- 
nuscule manuscripts of the New Testament outnumber the majuscule 
manuscripts by more than ten to one, and although one must make 
allowance for the greater antiquity of the majuscule style (and con- 
sequently the greater likelihood of the destruction of such manu- 
scripts through the ravages of time), very much of the disparity in the 
number of survivors must be due to the increased ease with which 
the minuscule copies could be produced. 

In times of economic depression, when the cost of vellum in- 
creased, the parchment of an older manuscript would be used over 
again. The original writing was scraped and washed off, the surface 
resmoothed, and the new literary material written on the salvaged ma- 
terial. Such a book was called a palimpsest ( which means "rescraped," 
from jzdXw and ipdco). One of the half-dozen or so most important 
parchment manuscripts of the New Testament is such a palimpsest; its 
name is Codex Ephraemi rescriptus. Written in the fifth century, it 
was erased in the twelfth century and many of the sheets rewritten 
with the text of a Greek translation of 38 treatises or sermons by 
St. Ephraem, a Syrian Church father of the fourth century. By applying 
certain chemical reagents and using an ultraviolet-ray lamp, scholars 

24. L. D. Reynolds and D. W. G. Wilson, Scribes and Scholars; A 
Guide to the Transmission of Greek and Latin Literature, 3rd ed. (Oxford, 
1991), p. 60. 

22 The Text of the New Testament 

have been able to read much of the almost obliterated underwriting, 
although the task of deciphering it is most trying to the eyes. 

In A.D. 692, the Council of Trullo (also called the Quinisext 
Council) issued a canon (no. 68) condemning the practice of using 
parchment from manuscripts of the Scriptures for other purposes. 
Despite the canon and the penalty of excommunication for one year, 
the practice must have continued, for of the 310 majuscule manu- 
scripts of the New Testament known today, 68 are palimpsests. 2,5 

Ancient scribes, when writing Greek, ordinarily left no spaces 
between words or sentences (this kind of writing is called scriptio 
continued, and until about the eighth century punctuation was used 
only sporadically. 26 At times, of course, the meaning of a sentence 
would be ambiguous because the division into words was uncertain. 
In English, for example, godisnowherk will be read with totally dif- 
ferent meanings by an atheist and by a theist ("God is nowhere" and 
"God is now here"). It must not be thought, however, that such am- 
biguities occur very often in Greek. 27 In that language it is the rule, 
with very few exceptions, that native words can end only in a vowel 
(or a diphthong) or in one of three consonants, v, q, and g. Further- 
more, it should not be supposed that scriptio continua presented 
exceptional difficulties in reading, for apparently it was customary in 

25. Besides Codex Ephraerni, they are the following: p c (= 024), p l,pr 
(= 025), Q (= 026), R (= 027), Z (= 035), 2 (= 040), 048, 062, 064, 065, 
066, 067, 068, 072, 078, 079, 086, 088, 093, 094, 096, 097, 098, 0103, 0104, 
0116, 0120, 0130, 0132, 0133, 0134, 0135, 0158, 0159, 0161, 0168, 0196, 0197, 
0208, 0209, 0225, 0229, 0233, 0240, 0245, 0246, 0247, 0248, 0249, 0250, 0254, 
0257, 0269, 0271, 0272, 0273, 0274, 0279, 0280, 0281, 0282, 0284, 0288, 0289. 
0297, 0306, 0307. For lists of majuscule palimpsests classified by successive 
centuries, see David C. Parker's "The Majuscule Manuscripts of the New 
Testament," in The Text of the New Testament in Contemporary Research, ed. 
by Ehrman and Holmes, pp. 24-5. 

26. Word division, however, is occasionally found in school and litur- 
gical texts, and scattered examples of punctuation, by point or spacing or a 
combination of both, are preserved in papyri from the third century B.C. 

27. Examples in the New Testament include the following. In Mark 
10.40, according to most editors, Jesus says "but it is for those for whom it 
has been prepared" {d/.X'oig rjioi/j.aoTai). This can also be read ahXoic; 
rJToifiaoTai, which means "it has been prepared for others." In Rom. 7,14, 
oida/iEV may be divided oidafiev. In 1 Tim. 3.16, xal dA.oyovfj.evwt; fieya 
eotlv may be taken as xai SfioXoyov/jtev wg/ieya iouv. 

The Making of Ancient Books 23 

antiquity to read aloud, even when one was alone. 28 Thus, despite 
the absence of spaces between words, by pronouncing to oneself 
what was read, syllable by syllable, one soon became used to read- 
ing scriptio continua. 29 

Christian scribes developed a system of contractions for certain 
"sacred" words. These nomina sacra, as they are called today, in- 
clude such frequently occurring nouns as Oeog, kvqioc,, r Irjoovg, 
XgioTog, and vide (which were contracted by writing only the first 
and last letters); jtvevfia, Aavid, oravgoc,, and ^rjrrjg (contracted 
by writing only the first two and the last letters); 7taxr\Q, y lagar\X, 
and aojrrjQ (of which the first and the last two letters were written); 

28. Besides scattered evidence from classical antiquity (collected by 
Josef Balogh, "Voces paginarum," Philologus, lxxxii [19271, pp. 84-109, 
202-31; also published separately), the statement in Acts 8.30 that Philip 
"heard" the Ethiopian treasurer reading from Isaiah the prophet implies 
that he had been reading aloud to himself. Compare this with the close of 
2 Maccabees: 

Here I will end my story. If it is well told and to the point, that is what I my- 
self desired; if it is poorly done and mediocre, that was the best I could do. 
For just as it is harmful to drink wine alone, or, again, to drink water alone, 
while wine mixed with water is sweet and delicious and enhances one's en- 
joyment, so also the style of the story delights the ears of those who read the 
work. (15.37-9) 

See also G. L. Hendrickson, "Ancient Reading," Classical Journal, xxv 
(1929), pp. 182-96; H. J. Chaytor, "The Medieval Reader and Textual 
Criticism," Bulletin of the John Rylands library, xxvi (1941-2), pp. 49-56; 
Eugene S. McCartney, "Notes on Reading and Praying Audibly," Classical 
Philology, lxiii (1948), pp. 184-7; B. M. W. Knox, "Silent Reading in Antiq- 
uity," Greek, Roman and Byzantine Studies, ix (1968), pp. 421-35; Paul 
Saenger, "Silent Reading: Its Impact on Late Medieval Script and Society," 
Viator. Medieval and Renaissance Studies, xiii (1982), pp. 369-414; P. J. 
Achtemeier, "Omne verhum sonat: The New Testament and the Oral 
Environment of Late Western Antiquity," Journal of Biblical Literature, cix 
(1990), pp. 3-27; K D. Gilliard, "More Silent Reading in Antiquity: non 
omne verbum sonabat," Journal of Biblical Literature, cxii (1993), 
pp. 689-94 (concludes: "the predominance of orality does not mean exclu- 
sivity, either in writing or in reading," p. 694); D. M. Schenkeveld, "Prose 
Usages of dxoveiv 'to Read,' " Classical Quarterly, xlii (1992), pp. 129-41. 

29. The experience of Hernias, who says he copied a little scroll of 
heavenly origin "letter by letter, for I could not make out the syllables" 

( Vision, II.i.4), suggests that the normal method of copying books was by 

24 The Text of the New Testament 

and avOgajnoz, 'IeQOVoakrjpL, and ovgavog (of which the first and 
last syllables were written). In order to draw the reader's attention to 
the presence of a nomen sacrum, the scribe would place a horizon- 
tal line above the contraction. 30 

In his article "Early Christian Book-Production" in volume 2 of 
The Cambridge History of the Bible, Skeat draws attention to the 
nomina sacra as being, together with the use of the codex, an indi- 
cation of "a degree of organization, of conscious planning, and uni- 
formity of practice among Christian communities which we have 
hitherto had little reason to suspect." 31 Where these patterns for 
Christian book production were first developed — Rome, Antioch, 
Alexandria, or Jerusalem — is discussed in detail by Roberts, 32 who in 
the end prefers Jerusalem. 

In the earlier ages of the Church, biblical manuscripts were pro- 
duced by individual Christians who wished to provide for them- 
selves or for local congregations copies of one or more books of the 
New Testament. Because the number of Christians increased rapidly 
during the first centuries, many additional copies of the Scriptures 
were sought by new converts and new churches. As a result, the 
speed of production sometimes outran the accuracy of execution. 
Furthermore, in preparing translations or versions for persons who 
knew no Greek, it occurred more than once (as Augustine com- 
plained) that "anyone who happened to gain possession of a Greek 

30. The standard work of Ludwig Traube, Nomina Sacra: Versuch 
einer Geschichte der christlichen Kurzung (Munich, 1907), is supplemented 
by the additional data collected by A. H. R. E. Paap, Nomina Sacra in the 
Greek Papyri of the First Five Centuries A.D.: The Sources and Some Deduc- 
tions (Leiden, 1959). Traube had fewer than 40 Greek papyri available; 
Paap cites evidence from 421 papyri of the first five centuries of the Christ- 
ian era. More recently, R. C. Nevius has published "Papyrus Witnesses to 
the Text of the 'Nomina Sacra' in the Apocalypse," in Akten des 21. Inter- 
nationalen Papyrologenkongresses, ed. by Barbel Kramer et al. (Stuttgart, 
1997), pp. 750-5. According to C. H. Roberts, the origin of the distinctively 
Christian custom of contracting the nomina sacra lies in the first century 
{[London] Times Literary Supplement, 10 March 1961, p. 160). See also 
idem, Manuscript, Society and Belief in Early Christian Egypt (London, 
1979), pp. 26-48; Gamble, Books and Readers, pp. 74-8; Kim Haines- 
Eitzen, Guardians of Letters: Literacy, Power, and the Transmitters of Early 
Christian Literature (New York, 2000), pp. 91-6. 

31. The West from the Fathers to the Reformation, ed. by G. W. H. 
Lampe (Cambridge, 1969), p. 73. 

32. Manuscript, Society and Belief, pp. 26-48. 

The Making of Ancient Books 25 

manuscript and who imagined that he had some facility in both Latin 
and Greek, however slight that might be, dared to make a transla- 
tion" (De doctrina Christiana, II.xi.l6). 

When, however, in the fourth century Christianity received offi- 
cial sanction from the state, it became more usual for commercial 
book manufacturers, or scriptoria, to produce copies of the books of 
the New Testament. 33 Sitting in the workroom of a scriptorium, sev- 
eral trained scribes, Christian and non-Christian, each equipped with 
parchment, pens, and ink, would write a copy of the book being 
reproduced as the reader, or lector, slowly read aloud the text of the 
exemplar. 34 In this way, as many copies could be produced simulta- 
neously as scribes were working in the scriptorium. It is easy to 
understand how in such a method of reproduction errors of tran- 
scription would almost inevitably occur. Sometimes the scribe would 
be momentarily inattentive or, because of a cough or other noise, 
would not clearly hear the lector. Furthermore, when the lector read 
aloud a word that could be spelled in different ways (e.g., in English, 
the words great and grate or there and their), the scribe would have 
to determine which word belonged in that particular context, and 
sometimes he wrote down the wrong word. (For examples of such 
mistakes, see pp. 254—5.) 

In order to ensure greater accuracy, books produced in scripto- 
ria were commonly checked over by a corrector (diOQOcorrjg) spe- 
cially trained to rectify mistakes in copying. His annotations in the 
manuscript can usually be detected today from differences in styles 
of handwriting or tints of ink. 

Scribes who were hired by a scriptorium to do a certain piece of 
work would be paid in accordance with the number of lines that they 
wrote. The standard length of line was originally a line of poetry, either 
a Homeric hexameter or an iambic trimeter. When prose works were 
copied, a line called a sticbos, having 16 (or sometimes 15) syllables, 
was frequently used as a measure for determining the market price 
of a manuscript. A price-fixing edict issued in A.D. 301 by Emperor 

33- Some have argued that in Alexandria a Christian scriptorium was 
in existence already by the latter half of the second century; see, e.g., 
G. Zuntz, The Text of the Epistles: A Disquisition upon the Corpus Paulinum 
(London, 1953), p. 273. A strong counterargument, that there is no evi- 
dence for Christian scriptoria before the fourth century, is made by Haines- 
Eitzen, Guardians of Letters, pp. 83-91. 

34. See especially T. C. Skeat, "The Use of Dictation in Ancient Book- 
Production," Proceedings of the British Academy, xlii (1956), pp. 179-208. 

26 The Text of the New Testament 

Diocletian set the wages of scribes at the rate of 25 denarii for 100 lines 
in writing of the first quality and 20 denarii for the same number of 
lines in writing of the second quality (what the difference was between 
the two qualities of writing is not mentioned). 35 According to the com- 
putation of Rendel Harris, the cost of producing one complete Bible, 
such as Codex Sinaiticus, would have come to about 30,000 denarii, a 
sizeable sum notwithstanding steadily rising inflation. 36 

The application of stichometric reckoning served also as a 
rough-and-ready check on the general accuracy of a manuscript, for 
obviously a document that was short of the total number of stichoi 
was a defective copy. On the other hand, such calculations were far 
from foolproof safeguards to the purity of the text, for only longer 
interpolations or omissions were likely to be disclosed by counting 
stichoi. In manuscripts of the Gospels that supply stichometric infor- 
mation, the most frequently appearing statistics are the round num- 
bers 2,600 for Matthew, 1,600 for Mark, 2,800 for Luke, and 2,300 for 
John. More precise figures, found in several manuscripts, are 2,560, 
1,616, 2,750, and 2,024, respectively, which imply, for example, the 
presence of 16.9-20 in Mark and the absence of 7.53-8.11 in John. 

Later, during the Byzantine period, copies of books were pro- 
duced by monks. In monasteries, there was much less pressure than 
in a commercial scriptorium to produce many copies at one time, so 
instead of writing at the dictation of a lector, individual monks, often 
working separately in their cells, would prepare copies of the Scrip- 
tures or other books either for themselves or for some benefactor of 
the monastery. Such a method of multiplying copies was not open to 
the same kinds of error involved in the dictation method, but another 
set of circumstances operated to make absolute accuracy difficult to 
secure. The act of copying entails four fundamental operations: 
(1.) reading to oneself (in antiquity no doubt reading half-aloud) a 
line or clause of the text to be copied, (2) retaining this material in 

35. De pretiis rerum venalium ("Concerning the Prices of Things 
Sold"), vii., 39 f., published in Corpus Inscriptionum Latuiarum, iii., 831; 
see E. R. Graser in Tenney Frank's An Economic Survey of Ancient Rome, v 
(Baltimore, 1940), p. 342. 

36. See New Testament Autographs, a supplement to the American 
Journal of Philology, 12 (Baltimore, 1882), p. 23. It is difficult to estimate 
the equivalent in modern currency. Some measure of comparison may be 
had, however, when it is known that in the preceding century under 
Caracalla (211-17) a legionary was paid a stipendium of 750 denarii per 
year in addition to his maintenance. 

The Making of Ancient Books 27 

one's memory, (3) dictating this material to oneself (either silently or 
half-aloud), and (4) the movement of the hand in executing the 
copy. Though several of these steps are executed almost simultane- 
ously, there was enough opportunity for the mind of a weary or half- 
awake scribe to play tricks that resulted in the most atrocious blun- 
ders (for examples, see pp. 258-9). 

Furthermore, prior to the invention of the fountain pen, it would 
have been necessary for the scribe to "re-ink" the pen by dipping it 
into an ink well. 37 The constant necessity to do so while copying 
documents provided the occasion for scribal distraction at the level 
of eye, memory, judgment, and pen. 38 

Besides various psychological causes of errors, physiological and 
external causes also conspired to make absolute accuracy extremely 
difficult to attain. One must bear in mind that the act of copying was 
in itself arduous and fatiguing, both because of the effort of sustained 
attention that it demanded as well as because of the cramped posi- 
tion of various muscles of the body. Though it seems strange to us 
today, in antiquity it was not customary to sit at a table or a desk 
while writing. Both literary 39 and artistic 40 evidence suggests that 

37. See Figure 1 1 for a notable example of a pattern of alternating darker 
and lighter letters, the result of re-inking the pen every four to six letters. 

38. See P. M Head and M. Warren, "Re-inking the Pen: Evidence from 
P.Oxy. 657 (p 13 ) Concerning Unintentional Scribal Errors," New Testament 
Studies, xliii (1997), pp. 466-73- 

39. An interesting piece of literary evidence is found in a colophon, 
dating from about the third century A.D., attached to a papyrus scroll con- 
taining portions of the third and fourth books of the Iliad {ad. by H. J. M. 
Milne, Catalogue of the Literary Papyri in the British Museum [London, 1927], 
pp. 21-2). According to the reading proposed by Wifstrand, the first two 
lines of the colophon are 'Eya> xoQwviq elfii ygafifidrcDV (pvXa^- xdlafiog/x 
eygaipe depict ^etg xai yovv; i.e., because the scribe held the sheet of pa- 
pyrus on his lap while he wrote, it could be said that the knee as well as the 
stylus and the right hand cooperated in producing what was written; see 
Albert Wifstrand, "Ein metrischer Kolophon in einem Homerus-papyrus," 
Hermes, Ixviii (1933), pp. 468-72. For further elaborations of Wifstrand's 
suggestion, see G. M. Parassoglou, "AEHIA XEIP KAI TONY. Some 
Thoughts on the Postures of the Ancient Greeks and Romans when Writing 
on Papyrus Rolls," Scrittura e Civilta, iii (1979), pp. 5-21; idem, "A Roll 
upon His Knees," Yale Classical Studies, xxxviii (1985), pp. 273-5. 

40. For a variety of artistic evidence bearing on the posture of scribes 
while writing, see the plates in A. M. Friend, Jr., "The Portraits of the 
Evangelists in Greek and Latin Manuscripts," Art Studies, v (1927), 

28 The Text of the New Testament 

Figure 4 Limestone statuette of Egyptian scribe, in the Louvre, P --■• ;. 

until the early Middle Ages it was customary for scribes either to 
stand (while making relatively brief notes) or to sit on a stool or 
bench (or even on the ground), holding their scroll or codex on their 
knees (see Fig. 4). 41 It goes without saying that such a posture was 

pp. 115-47, and vii (1929), pp. 3-29; W. H. P. Hatch, Greek and Syrian 
Miniatures in Jerusalem (Cambridge, MA, 1931); and B. M. Metzger. "When 
Did Scribes Begin to Use Writing Desks?" in his Historical and Literary 
Studies, Pagan, Jewish, and Christian (Grand Rapids, MI, 1968), f.L, 
For a discussion of a miniature in the Rossano Gospels depicting the tnai 
of Jesus (see Fig. 14, p. 85), which suggests that it was customary for 
stenographers in a law court to stand while writing even though a ta^ii 
was available, see William C. Loerke, "The Miniatures of the Trial ir the 
Rossano Gospels," Art Bulletin, xliii (1961), pp. 171-95. 

41. The so-called writing tables found at Qumran, which re- 
built up by the archaeologists to the height of tables today, were oi giiu Ih 
but 20 inches high, too low to serve as writing desks; see the discussion of 
Bruce M. Metzger, "The Furniture of the Scriptorium at Qumran," Revue de 
Qumran, i (1958-9), pp. 509-15. 

The Making of Ancient Books 29 

more tiring than sitting at a desk or writing table, though the latter 
must have been tiring enough to scribes thus occupied 6 hours a 
day 42 month after month. 

Something of the drudgery of copying manuscripts can be 
learned from the colophons, or notes, which scribes not infrequently 
placed at the close of their books. A typical colophon found in many 
nonbiblical manuscripts reveals in no uncertain terms what every 
scribe experienced: "He who does not know how to write supposes 
it to be no labor; but though only three fingers write, the whole body 
labors." A traditional formula appearing at the close of many manu- 
scripts describes the physiological effects of prolonged labor 
at copying: "Writing bows one's back, thrusts the ribs into one's 
stomach, and fosters a general debility of the body." In an Armenian 
manuscript of the Gospels, a colophon complains that a heavy snow- 
storm was raging outside and that the scribe's ink froze, his hand 
became numb, and the pen fell from his fingers! It is not surprising 
that a frequently recurring colophon in manuscripts of many kinds is 
the following comparison: "As travellers rejoice to see their home 
country, so also is the end of a book to those who toil [in writing]." 
Other manuscripts close with an expression of gratitude: "The end of 
the book; thanks be to God!" 

Upon more mature reflection, however, many scribes doubtless 
judged the rewards of copying the Scriptures to outweigh the dis- 
comforts they experienced during the long hours of writing. Thus, 
Cassiodorus, that remarkable rhetorician-philosopher and prime 
minister to the Ostrogothic princes of Italy, who later became a monk 
and founded the monastery of Vivarium, noted for its school of Latin 
paleography, dilates upon the spiritual compensations gained by the 
faithful scribe: 

By reading the divine Scriptures [the scribe] wholesomely instructs his 
own mind, and by copying the precepts of the Lord he spreads them 
far and wide. What happy application, what praiseworthy industry, to 
preach unto people by means of the hand, to untie the tongue by 
means of the fingers, to bring quiet salvation to mortals, and to fight the 
Devil's insidious wiles with pen and ink! For every word of the Lord 

42. See the anonymous complaint, dating perhaps from the ninth 
century, quoted by Falconer Madan in his Books in Manuscript (London, 
1893), p. 37: "Ardua scriptorum prae cunctis artibus ars est: | Difficilis 
labor est, durus quoque flectere colla, | Et membranas bis ternas sulcare 
per floras." 

30 Tiik Text of the New Testament 

written by the scribe is a wound inflicted on Satan. And so, though 
seated in one spot, the scribe traverses diverse lands through the dis- 
semination of what he has written. . . . Man multiplies the heavenly 
words, and in a certain metaphorical sense, if I may dare so to speak, 
three fingers are made to express the utterances of the Holy Trinity. O 
sight glorious to those who contemplate it carefully! The fast-travelling 
reed-pen writes down the holy words and thus avenges the malice of 
the Wicked One, who caused a reed to be used to smite the head of 
the Lord during his Passion.' 13 

In view of the difficulties involved in transcribing ancient books, 
it is the more remarkable how high was the degree of achievement 
of most scribes. The fact is that in most manuscripts the size of the 
letters and the ductus of the script remain surprisingly uniform 
throughout even a lengthy document. 

In order to secure a high degree of efficiency and accuracy, cer- 
tain rules pertaining to the work of scribes were developed and en- 
forced in monastic scriptoria. The following are examples of such 
regulations prepared for the renowned monastery of the Studium 
at Constantinople. About A.D. 800 the abbot of this monastery, 
Theodore the Studite, who was himself highly skilled in writing an 
elegant Greek hand, included in his rules for the monastery severe 
punishments for monks who were not careful in copying manu- 
scripts. 44 A diet of bread and water was the penalty set for the scribe 
who became so much interested in the subject matter of what he was 
copying that he neglected his task of copying. Monks had to keep 
their parchment leaves neat and clean, on penalty of 130 penances. 
If anyone should take without permission another's quaternion (that 
is, the ruled and folded sheets of parchment), 50 penances were pre- 
scribed. If anyone should make more glue than he could use at one 
time and it should harden, he would have to do 50 penances. If a 

43. Cassiodori Senatoris Instituliones, edited from the manuscripts by 
R. A. B. Mynors (Oxford, 1937), 1. xxx. 1. For perceptive comments on 
how Christians approached their manuscripts, how they produced them, 
and how they conceived of their use, see Claudia Rapp's discussion, 
"Christians and Their Manuscripts in the Greek East in the Fourth Century." 
in Scritture, libri e testi rielle aree provinciali di Bisanzio, ed. by Guglielmo 
Cavallo et al. (Spoleto, 1991), pp. 127^8. 

44. The text of these rules is in Migne, Patrologia Graeca, xcix, 
pp. 1739 f. For a competent biography of Theodore, see Alice Gardner's 
Ibeodore of Studium, His Life and Times (London, 1905). 

The Making of Ancient Books 31 

scribe broke his pen in a fit of temper (perhaps after having made 
some accidental blunder near the close of an otherwise perfectly 
copied sheet), he would have to do 30 penances. 

Added Note on Colophons 

Besides the several colophons quoted above, which directly or 
indirectly witness to the difficulties involved in transcribing a book, 
there are many other kinds of colophon. Some provide the name of 
the scribe and occasionally the place and date of writing. Obviously, 
such information is of the greatest value to the paleographer in trac- 
ing the background and family relationships of manuscripts. 45 

Some colophons are in the form of a blessing or prayer or an 
invitation to the reader to offer such a prayer; e.g., "Whoever says, 
'God bless the soul of the scribe,' God will bless his soul." The 
following prayer is found at the close of a Psalter copied in the 
year 862: 

IXeoc, rq> ygaipavTi, xvqle, 
ao(pia Tolg avayivwoxovoi, 
XO.QI? Totg axovovoi, 
owTtjgia role; xexTn/xevoig- ajuqv. 

(Mercy be to him who wrote, O Lord, wisdom to those who 
read, grace to those who hear, salvation to those who own [this 
codexl. Amen.) 

An extended prayer at the end of a Coptic-Arabic manuscript of 
the Gospels contains the following: 

O reader, in spiritual love forgive me, and pardon the daring of him 
who wrote, and turn his errors into some mystic good. . . . There is no 
scribe who will not pass away, but what his hands have written will re- 
main for ever. Write nothing with your hand but that which you will be 

45. For a list of the names of Greek scribes, see Marie Vogel and 
Viktor Gardthausen, Die griechiscben Schreiber des Mittelalters und der 
Renaissance (Leipzig, 1909). According to Christopher de Hamel, the 
Benedictine monks of Le Bouveret in Switzerland have been publishing 
their vast index of signed colophons (more than 19,000) in medieval man- 
uscripts of various kinds, where "some scribes sign themselves Johannes, 
or Rogerius, which tells us very little. A gratifyingly large number are 
women, which one might not have expected" {Medieval Craftsmen: Scribes 
and Illuminators [Toronto, 1992], p. 43). 

32 Thk Text of the New Testament 

pleased to see at the resurrection. . . . May the Lord God Jesus Christ 
cause this holy copy to avail for the saving of the soul of the wretched 
man who wrote it. 46 

In some manuscripts, one finds curse colophons, which were in- 
tended to serve as the equivalent of modern theft insurance. For ex- 
ample, in a twelfth-century Greek lectionary of the Gospels, now in 
the library of Princeton Theological Seminary (Fig. 3), there is a 
colophon stating that the volume was donated to the church of 
St. Saba at Alexandria: "No one therefore has authority from God to 
take it away under any condition, and whoever transgresses this will 
be under the wrath of the eternal Word of God, whose power is great. 
Gregory, Patriarch by the grace of God, wrote this." 47 

Less formal are the conversational jottings that occasionally 
stand at the close of a manuscript or in the margins of folios through- 
out a document. Though scribes were forbidden to talk to one an- 
other in the scriptorium, the more irrepressible found devious ways 
to communicate with each other. One such means was to jot remarks 
on the margin of the page being transcribed and to show it to one's 
neighbor. The margins of a ninth-century Latin manuscript of 
Cassiodorus' commentary on the Psalms contain a variety of com- 
monplace remarks written in Irish. For example: "It is cold today." 
"That is natural; it is winter." "The lamp gives a bad light." "It is time 
for us to begin to do some work." "Well, this vellum is certainly 
heavy!" "Well, I call this vellum thin!" "I feel quite dull today; I don't 
know what's wrong with me." 48 

How did it happen that the head of the scriptorium allowed his 
monks to disfigure a manuscript with such trivialities? One may per- 
haps conjecture that the manuscript was written in a continental 

46- [George Horner,] The Coptic Version of the New Testament in the 
Northern Dialect, otherwise called Memphitic and Bohairic, i (Oxford, 
1898), pp. cxlvi f. 

47. For other curse colophons, see Lawrence S. Thomson, "A Cursory 
Survey of Maledictions," Bulletin of the New York Public Library, lvi (1952), 
pp. 55-74. 

48. The Irish text, with a German translation, is given by Kuno Meyer, 
"Neu aufgefundene altirische Glossen," Zeitschrift fiir celtische Philologie, 
viii (1912), pp. 173-7. For many other colophons and notes in Greek and 
Latin manuscripts, see W. Wattenbach, Das Schriftwesen im Mittelalter, 3te 
Aufl. (Leipzig, 1896), pp. 491-534; Gardthausen, Griechische Palaeogra- 
phie, pp. 424 ff. 

The Making of Ancient Books 33 

monastery, where the authorities knew no Irish and therefore the 
scribes from Ireland felt they could play pranks with impunity. When 
asked what he had written, the scribe might point to some pious sen- 
tences in Latin in the top margins of preceding pages and say 
"Merely the Irish equivalents of sentences like these!" 49 

In order to ensure accuracy in transcription, authors would 
sometimes add at the close of their literary works an adjuration di- 
rected to future copyists. So, for example, Irenaeus attached to the 
close of his treatise On the Ogdoad the following note: 

I adjure you who shall copy out this book, by our Lord Jesus Christ and 
by his glorious advent when he comes to judge the living and the dead, 
that you compare what you transcribe, and correct it carefully against 
this manuscript from which you copy; and also that you transcribe this 
adjuration and insert it in the copy. 50 

IV. "Helps for Readers" in New 
Testament Manuscripts 

Many manuscripts of the New Testament are provided with a va- 
riety of what may be called "helps for readers," that is, aids to assist 
in the private as well as public reading of the Scriptures. This mate- 
rial originated in various places and at various times. It was handed 
on from generation to generation, and, as would be expected, it 
grew in volume with the passage of time. The following is an enu- 
meration of some of these aids found in Greek manuscripts. 51 

49. For other instances of conversation written in the margins of 
manuscripts, see W. M. Lindsay, Early Irish Minuscule Script (Oxford, 
1910), p. 42. 

50. Apud Eusebius, Hist. Eccl. v.xx.2. Cf. the warning given in 

Rev. 22.18, and see W. C. van Unnik, "De la regie Mr/re TiQoaQelvai fiijre 
dcfreXeiv dans l'histoire du canon," Vigiliae Christianae, iii (1949), pp- 1-36. 

51. For a discussion of the accessories in Latin manuscripts of the 
Bible, see [Donatien De Bruyne,] Sommaires, divisions et rubriques de la 
Bible latine (Namur, 1914); idem, Prefaces de la Bible latine (Namur, 1920); 
Patrick McGurk, Latin Gospel Books from A. D. 400 to A. D. SCO (Brussels, 
1961). Jiirgen Regul, having utilized the rich resources of the Vetus Latina 
Institut, has considerably broadened the Latin textual basis for the presence 
of Gospel prologues, citing about 70 manuscripts in his apparatus. 
Contrary to De Bruyne, Regul concludes that the so-called anti-Marcionite 

34 Ti if. Text of the New Testament 

1. Chapter Divisions (X£(pdAaia) 

The oldest system of capitulation that is known to us is that pre- 
served in the margins of Codex Vaticanus (B) of the fourth century. 
In this manuscript, there are 170 sections in Matthew, 62 in Mark, 
152 in Luke, and 50 in John. Another system of chapter divisions 
is found in Codex Alexandrinus (A) of the fifth century as well as 
in most other Greek manuscripts. According to this capitulation, in 
Matthew there are 68 xecpdXaia, in Mark 48, in Luke 83, and in John 
18. In no case does the first chapter stand at the beginning of a 
book, probably because of the custom of scribes to refer to the 
opening section of a book as the ngooifiiov, or preface. Thus, xeip. 
a of Mark begins at Mark 1.23. 

For the Book of Acts, several systems of chapter division are cur- 
rent in the manuscripts. Codex Vaticanus has two sets of capitula- 
tion, one of 36 chapters and the other of 69 chapters. According to 
Hatch, the chapter numbers of the former system were inserted in 
the margin of the manuscript by a very early hand — perhaps by the 
diogdwrrfg or possibly by the scribe himself — and the other system 
of chapter numbers was added somewhat later by another scribe. 52 
In Codex Sinaiticus of the fourth century, the system of 69 chapters 
was added by someone to the first part of Acts (chapters 1-15), but 
for some unknown reason it was not continued to the end of the 

Most other Greek manuscripts of Acts have a system of 40 
x£<pdlaia. In some manuscripts, the division of Acts into sections 
was carried still further, 24 of the 40 chapters being subdivided into 
smaller sections (vJiodiaigeoeig). There were 48 such smaller subdi- 
visions, making a total of 88 xefyakata and vnodiaiQeaeig. It was 
inevitable that the distinction between the larger and smaller sec- 
tions would be confused, and in some manuscripts they are num- 
bered consecutively throughout the book. 

Both the Pauline and the Catholic Epistles were also divided 
into chapters, and many of these were subdivided into smaller 

prologues do not come from one hand, nor are they anti-Marcionite in 
character or origin. Nevertheless, he felt that he must give his monograph 
the title Die Antimarcionitischen Evangelienprologe (Aus der Gescbichte 
der Lateinischen Bibel 6; Freiburg, 1969). 

52. W. H. P. Hatch, "Euthalius," in The Twentieth Century Encyclope- 
dia of Religious Knowledge, i (Grand Rapids, MI, 1955), p. 400. 

The Making of Ancient Books 



-M»Ii*i»d^ oi'er * o uUiy6- 

1 .;"< -?• 4^*^ ' AJX ' **>* "^ 
cLsrovrro enotLfa^i^tstiV 

aJTDDAuorcM »i'^i(l v • I 
$*• 'i o u$*m o t # gli-p «u/ <yuu 

93«VA^v B V" T ^'*6 OL l' • 
•Tol> T T^»l<cL^^raXl/er^f<o , otu 

p e a" •': 1-amjr o uxlo{ X 6 */ 

<ra^>"TDu«Toy»TDyX<'V H. 

K^ySy €31" n»H I yi&e 

*wo» <( u'm « m oy\ey>tiir 
f ° V X rw> <rrpa» *ro \i e 

■ I 

of oy orpa>ai»y/CLu«TD|)».« t» 

X € vC*" *** • ** ©•■lA*' 
nocr« , n»yu*uHX6a/£>, ' 

ovti^eoy^dry uaoiX&cL/ , 



I'uuiuif gTffy«T*£rft 

i'»; xwp dcnui*TwP ot ff* 
jX*V» y I Uiuif«T»t> 

/ v ' * 

r piau6y«ra>youD 4 


• 2 US 


Figure 5 Greek Gospel Lectionary 562 (written, according to a colophon, in 
the city of Capua, A.D. 991), in the Vatican Library, Rome; John 19.10-16 and 
Matt. 27.3-5 (see pp. 46-7). Actual size 9^ X 1% inches. 

36 The Text of the New Testament 

sections. 53 Codex Vaticanus has two sets of chapter divisions for the 
Epistles, an earlier and a later system. In the Pauline Epistles, the 
earlier division enumerates the chapters consecutively throughout 
the corpus (for the clue that this provides regarding an ancestor of 
Vaticanus, see p. 69, n. 27). 

The Book of Revelation was supplied with a highly artificial sys- 
tem of divisions. In the latter part of the sixth century, Archbishop 
Andrew of Caesarea in Cappadocia wrote a commentary on the 
book that gives a "spiritual" exegesis. Instead of asking what mater- 
ial was in the book and into how many parts it could most appro- 
priately be divided, he divided the book into 24 Xoyoi, or discourses, 
because of the 24 elders sitting on thrones about the throne of God 
(Rev. 4.4). He further reflected that the nature of each of the 24 
elders was tripartite (ocdfia, ipvxi], and Jtvevfid) and, therefore, di- 
vided each loyog into three xe<j>dXaia, making a total of 72 chapters 
for the book. 

2. Titles of Chapters (jlxXoi) 

Each of the Kzfyakaia of the system found in codex Alexandri- 
nus and in most other later manuscripts is provided with a rkXog. 
This is a summary heading placed in the margin and describing the 
contents of the chapter (Fig. 6). These titles customarily begin with 
the word about or concerning and are not infrequently written with 
red ink. Thus, HE(f>. a' of John, which begins at 2.1, has the title "Con- 
cerning the Marriage at Cana" (jieqi tov ev Kava ydfiov). 54 All of the 
titXol for one book are frequently listed and placed before that book 
as a summary outline of what follows. 

53. For information regarding the number of each of these divisions in 
each of the epistles, see W. H. P. Hatch, Facsimiles and Descriptions of 
Minuscule Manuscripts of the New Testament (Cambridge, MA, 1951), p. 25. 
See also H. K. McArthur, "The Earliest Divisions of the Gospels," in Studia 
Evangelica, iii, ed. by F L. Cross (Texte und Untersuchungen, lxxxviii; 
Berlin, 1964), pp. 266-72. 

54. For a complete list of xhloi, see H. von Soden, Die Schriften des 
Neuen Testaments in ihrer altesten erreichbaren Textgestalt, l.i (Berlin, 
1902), pp. 405 ff. 

The Making of Ancient Books 


t .J 

n'«ii'iein.droif.Gi«A! A'potic^' 

'i-miifpAKAT 1'c ^svAiiiC-pinATfr 

in aa° bi AI« "i*0.6*ri oy oyci Ati f 

VHoyVvroYAVmY. 4<|>ioiiaio - 

iiinici'iir.iik.iip'ri.n*. • \uroi*ru> 

iiApAAY' 1 '"** 1 *' coi \i- i'u.»-tii t fr 
p0.l«AiX'i>t>M»ro»* ifpxiUT'PO'' 


SOyiCAl yitarOQit*. tomdi KO' 1 
Coy* lCAIlil'OpOlia\oCHOO. If^ 
A DAOU'OMUpAliA'r Dll , 6y»lA6f' * 
OKAIlTIOtlll AM 'IMUtt • (Jud'O 

oy i CrACOAin aii TACiatAo 
y*VOI M'l'Oii OH*Ol'OH 'tACO^ 
O YAPtl O 'i* O 0\ *V Ui I'-Gl AO M O I !•'%* 
ft AC O. A II. IC A I It AC. 6 6 VAOCHp 

^O'TOlipOCA'y^rou.ItfAldA.IAA . 

1*7 U> M M> GlAGN *C\$Vm "f OH 

• • M * * 

C t* AC ,tt K O AU Y ^ l« CO^Idf 





Figure 6 Codex Basiliensis (E, eighth century), University Library, Basle; 
Mark 2.9-14 (with tittos, lectionary equipment, and Eusebian canon numer- 
als; see pp. 36, 38, and 46). Actual size 9 X 6!4 inches. 

38 The Text of the New Testament 

3. Eusebian Canons 

An ingenious system was devised by Eusebius of Caesarea to aid 
one in locating parallel passages in the Gospels. 55 Apparently, his 
system was found to be highly useful, for it appears in a great 
number of manuscripts of the Gospels in Greek as well as in Latin, 
Syriac, Coptic, Gothic, Armenian, and other versions. 

The synopsis or harmony was prepared as follows. Each Gospel 
was divided into longer or shorter sections, depending on the rela- 
tion of each section to one or more parallels in the other Gospels. 
These sections were numbered consecutively throughout each 
Gospel (there are 355 in Matthew, 233 in Mark, 342 in Luke, and 232 
in John). Then, Eusebius prepared ten tables or canons (xavoveg), 
the first containing references by numerals to parallel passages 
found in all four Gospels; the second to passages common to 
Matthew, Mark, and Luke; the third to passages common to Matthew, 
Luke, and John; and so on, until almost all possible combinations of 
Gospels were exhausted. 56 The final table gives references to matter 
peculiar to each Gospel alone. These tables of numerals, written out 
in columns, customarily occupy the opening pages of Gospel manu- 
scripts. 57 Then, in the margin of the Gospel text of the manuscript, 
alongside or beneath the numeral of the consecutive sections, was 
written the numeral of the canon table in which that section could be 
found. For example, in the Gospel according to John one comes 
upon the statement "For Jesus himself testified that a prophet 
has no honor in his own country" (4.44); whoever .wishes to consult 
the parallels will find in the margin the numerals -■ (=^> By turn- 
ing to the first canon table and running an eye down the column of 

55. See C. Nordenfalk, "The Eusebian Canon-Tables: Some Textual 
Problems," Journal of 'Theological Studies, xxxv (1984), pp. 96-104. 

56. There is no table of references to sections in Mark, Luke, and John 
or to sections in Mark and John. 

57. The bare lists of numerals soon attracted the attention of artists, 
who ornamented the spaces between with decorative columns, arches, 
architraves, birds, flowers, etc.; see Fig. 20, p. 116 and Carl Nordenfalk, 
Die spdtantiken Kanontafeln: Kunstgeschicbtliche Studien iXber die Euse- 
bianiscbe Evangelien-Konkordanz in den vier ersten Jahrhunderten ibrer 
Geschichte, 2 vols. (Goteborg, 1938). For additional investigation, see idem, 
"The Apostolic Canon Tables," Gazette des beaux-arts, Ixii (1963), 

pp. 17-34; H. K. McArthur, "The Eusebian Sections and Canons," Catholic 
Biblical Quarterly, xxvii (1965), pp. 250-6. 

The Making of Ancient Books 39 

numerals referring to sections in John, one finds 35. In the horizontal 
line opposite this numeral is found the numeral 142 standing in the 
column of sections of Matthew, 51 in Mark, and 21 in Luke. Since, as 
was said above, the sections in each Gospel are numbered consecu- 
tively, it is an easy matter to find in the other three Gospels each of 
the sections that contains the parallels to the statement in John. 

For the added convenience of the user, in some manuscripts the 
numerals referring to the sections that are parallel to the passage on 
any given page are provided in the lower margin of the page (see 
Fig. 6, p. 37, and Fig. 20, p. 116) so that one can turn at once to the 
appropriate sections in the other Gospels. 

Eusebius explained his ingenious system in a letter to a fellow 
Christian named Carpianus, and a copy of this letter appears with the 
canon tables on the opening folios of many Gospel manuscripts. 58 
Several twentieth-century editions of the Greek Testament are pro- 
vided with the Eusebian canon tables and numbers, which thus con- 
tinue to be of service to present-day readers of the Gospels. 

4. Hypotheses, Bioi, Euthalian Apparatus 

The hypothesis (vjioOemg, Latin argumentum) is a prologue or 
brief introduction to a book, supplying the reader with a certain 
amount of information concerning the author, content, and circum- 
stances of composition of the particular book. The form and content 
of such hypotheses are often conventional and stereotyped. In some 
manuscripts, the hypotheses for the Gospels are ascribed to Eusebius, 
but more often they are anonymous. Anti-Marcionite Gospel pro- 
logues are extant in Latin manuscripts from the fifth to the tenth 
centuries. The Marcionite prologues to the (ten) Pauline Epistles were 
taken over practically unaltered by the Roman Catholic Church for the 
Latin Vulgate. 59 

A longer statement of traditional information concerning the life 
of each evangelist (called his fiiog) sometimes appears with the 

58. An English translation of Eusebius' letter was published by Harold 
H. Oliver in Novum Testamentum, iii (1959), pp. 138-45. 

59. For a survey of investigations concerning the Marcionite pro- 
logues, see B. M. Metzger in The Text, Canon, and Principal Versions of 
the Bible, ed. by E. E. Flack and B. M. Metzger (Grand Rapids, MI, 1956), 
pp. 24-6, supplemented by J. Regul's research on the anti-Marcionite 
prologues (see p. 33, n. 51). 

40 The Text of the New Testament 

hypothesis. The lives are attributed to an otherwise unknown 
Dorotheus of Tyre or to Sophronius, the patriarch of Jerusalem in the 
first half of the seventh century. 

Several different prologues that define the word gospel and pro- 
vide general information about all four Gospels collectively occur in 
various manuscripts. Besides lists of the 12 apostles, the traditional 
names of the 70 (or 72) disciples in Luke 10.1 ff. are given on the 
authority of Dorotheus and Epiphanius. 

For the Book of Acts, several different hypotheses are current; 
some are anonymous and others are taken from Chrysostom's com- 
mentary and homilies on Acts. For Acts and the Epistles, a consider- 
able apparatus of auxiliary materials circulated under the name of 
Euthalius or Evagrius. 60 Besides chapter divisions and hypotheses, 
these included a lengthy sketch of the life, writings, and chronology 
of the apostle Paul; a brief statement of the martyrdom of Paul; a 
table of Old Testament quotations in the Epistles; a list of places at 
which the Epistles were thought to be written; and a list of the names 
associated with Paul's in the headings to the Epistles. How much of 
this supplementary material was drawn up by Euthalius and how 
much was added later is not known. 

5. Superscriptions and Subscriptions 

In the oldest manuscripts of the New Testament, the titles of 
the several books are short and simple, for example, KATA 
MA0GAION or nPOZ PQMAIOYX In later centuries, these titles 
became longer and more complex (see p. 270). 

60. Scholarly opinion as to who this Euthalius or Evagrius was and 
when he lived has vacillated widely; see, for example, J. A. Robinson, 
Euthaliana (Cambridge, 1895); E. von Dobschiitz, "Euthaliusstudien," 
Zeitschrift fur Kirchengeschichte, xxix (1899), pp. 107-54; von Soden, op. 
cit., pp. 637-82; C. H. Turner in Hastings' Dictionary of the Bible, extra 
vol. (1904), pp. 524-9; G. Bardy in Supplement au Dictionnaire de la Bible, 
ii (1934), cols. 1215-18; G. Zuntz, "Euthalius = Euzoius?" Vigiliae Chris- 
tianize, vii (1953), pp. 16-22; J. W. Marchand, "The Gothic Evidence for 
'Euthalian' Matter," Harvard Theological Review, xli (1956), pp. 159-67; 
Sebastian Brock, "The Syriac Euthalian Material and the Philoxenian 
Version of the New Testament," Oriens Christianus, lxviii (1984), 
pp. 170-95; Hatch; and S. Dopp and W. Geerlings, eds., Dictionary of 
Early Christian Literature (New York, 2000), pp. 222-3. 

The Making of Ancient Books 41 

The subscriptions appended to the end of the books were orig- 
inally (like the titles) brief and simple, merely indicating the close of 
the book. As time passed, these became more elaborate and often in- 
cluded traditional information regarding the place at which the book 
was thought to be written and sometimes the name of the amanuen- 
sis. The King James Version includes the subscriptions to the Pauline 

6. Punctuation 

As was mentioned above, the earliest manuscripts have very lit- 
tle punctuation. The Chester Beatty papyri and the Bodmer papyri 
(Figs. 7, 8) have only an occasional mark of punctuation, 61 as do the 
early majuscule manuscripts. A diaeresis is sometimes placed over an 
initial iota or upsilon. During the sixth and seventh centuries, scribes 
began to use punctuation marks more liberally, though the sign of 
interrogation is rarely found before the ninth century. Gradually, 
the earlier sporadic and somewhat haphazard usage gave way to a 
fuller and more or less consistent style. In addition to the usual 
marks of punctuation, a syllable divider was often inserted after non- 
Greek proper names, as in such sequences as that in Matt. 1.2, 
ABPAAM'ErENNHSEN, where it stands to warn the reader 
against dividing it a-braa-me-gen-ne-sen. 

7. Glosses, Scholia, Commentaries, 
Catenae, Onomastica 

Glosses are brief explanations of difficult words or phrases. They 
were usually written in the margin of manuscripts, though occasion- 
ally they were placed between the lines. In the latter case, a Greek 
manuscript might be glossed throughout with a Latin interlinear, and 
a Latin manuscript with an Anglo-Saxon interlinear. 

Scholia are interpretive remarks of a teacher placed beside the 
text in order to instruct the reader. When scholia are systematically 
developed to elucidate continuously the entire text, rather than 
being merely random notes on certain passages, the work is called a 
commentary. Scholia and commentaries are sometimes placed in 
the margins around the Scripture text and sometimes interspersed 

6l. For a list of papyri that have marks of punctuation, see Guilelmus 
Flock, "De graecorum interpunctionibus" (Diss., Bonn, 1908), pp. 14 ff. 

42 Tl. >f the New Testament 




rr^f t» WittT« *v.-Hfc* rt1<|gh**tft ;m"^^5* 

Figure 7 Chestci leatty Biblical r_:pyrus II (t» 4f \ third century), Chester 
Beatty i seum, Dublin, Rom. (5.29-33, 16.25-7, 1-3 (see pp. 54-5). Actual 
size 9 X S inches. 

The Making of Ancient Books 43 

between sections of the Scripture text (see Fig. 12, p. 78). In majus- 
cule manuscripts, the explanatory materials are usually in minuscule 
script, though in codex Zacynthius (S) of the seventh century, the 
commentary is written in majuscules. When the text is in minuscules, 
occasionally the scholia are in small majuscules. 

Catenae are literally "chains" of comments extracted from older 
ecclesiastical writers. The identity of the original commentator is in- 
dicated by prefixing the abbreviation of his name, though through 
carelessness this mark of identification is sometimes missing. 

Onomastica are philological aids that purport to give the mean- 
ing and etymology of proper names. Like so much of ancient ety- 
mological lore — preserved, for example, throughout the works of 
Philo and in Plato's Cratylus — these explanations almost always are 
arbitrary and fanciful. 

8. Artistic Adornment 

In addition to an ornamental headpiece at the beginning of a 
book and illuminated initials, the Greeks of the Byzantine period 
sought to assist the reader in understanding the significance of the 
Scriptures by providing pictures of various kinds. 62 Some are por- 
traits of Christ and of his apostles, and others depict scenes or events 
that are narrated in the New Testament (see Fig. 14, p. 85)- The por- 
traits of the Evangelists fall into two main classes, those in which the 
figures are standing and those in which they are sitting. From a com- 
parison with Hellenistic representations of ancient Greek poets and 
philosophers, it appears that Christian artists, who had no knowl- 
edge of the likenesses of the Evangelists, adopted and adapted 
familiar portraits of pagan authors in contemporary art. According to 

62. For reproductions of representative miniatures, see W. H. P. Hatch, 
Greek and Syrian Miniatures in Jerusalem (Cambridge, MA, 1931); E. J. 
Goodspeed, D. W. Riddle, and H. R. Willoughby, The Rockefeller 
McCormick New Testament (Chicago, 1932); E. C. Colwell and H. R. 
Willoughby, The Four Gospels of Karahissar (Chicago, 1936); H. R. 
Willoughby and E. C. Colwell, The Elizabeth Day McCormick Apocalypse 
(Chicago, 1940); Carl Nordenfalk, Die spdtantiken Zierbuchshaben 
(Stockholm, 1970); Kurt Weitzmann, Late Antique and Early Christian Book 
Illumination (New York, 1977); J. J. G. Alexander, The Decorated Letter 
(London, 1978); Carl Nordenfalk, Early Book Illumination, 2nd ed. 
(Stockholm, 1988); J. J. G. Alexander, Medieval Illuminators and Their 

44 The Text of the New Testament 

the investigations of A. M. Friend, Jr., 63 all the early Christian portraits 
of the Evangelists go back to two main sets of four portraits each: one 
set was of the four philosophers Plato, Aristotle, Zeno, and Epicurus 
and the other set was of the four playwrights Euripides, Sophocles, 
Aristophanes, and Menander. 

The earliest New Testament manuscripts that contain miniatures 
are two deluxe copies on purple vellum of the sixth century, Codex 
Sinopensis (O) and Codex Rossanensis CT, see pp. 79 and 85). In the 
course of time, custom and tradition came to dictate the proper form 
and colors that artists should use in painting the several cycles of 
biblical scenes and characters. These traditional directives are given 
in the Byzantine Painters' Manual compiled by Dionysius, a monk 
of Fourna d'Agrapha. 64 Unfortunately, New Testament scholars have 
not yet availed themselves of the evidence supplied by artistic 
adornment in the investigation of family relationships of Byzantine 
manuscripts. 65 

Methods (London, 1994); Christopher de Hamel, Scribes and Illuminators 
(London, 1992); idem, A History of Illuminated Manuscripts, 2nd ed. 
(London, 1994); idem, The British Library Guide to Manuscript Illumina- 
tion, History and Techniques (London, 2001). 

63- "The Portraits of the Evangelists in Greek and Latin Manuscripts," 
Art Studies, v (1927), pp. 115-46, and vii (1929), pp. 3-29. Friend's con- 
clusions were supported and expanded by later investigators, such as Kurt 
Weitzmann, in Die byzantinische Buchmalerei des 9. und lO.Jahrhunderts 
(Berlin, 1935), pp. 23-46; D. T. Tselos, "Unique Portraits of the Evangelists 
in an English Gospel-book of the Twelfth Century," Art Bulletin, xxxiv 
(1952), pp. 257-77; E. Rosenbaum, "Evangelist Portraits of the Ada School 
and Their Models," Art Bulletin, xxxviii (1956), pp. 81-90; K. Weitzmann, 
"Book Illustration of the Fourth Century: Tradition and Innovation," in 
Studies in Classical and Byzantine Manuscript Illumination, ed. by 
Herbert L. Kessler (Chicago, 1971), pp. 112-13, 115; Joyce Kibiski, "The 
Medieval 'Home Office': Evangelist Portraits in the Mount Athos Gospel 
Book, Stravronika Monastery, Ms. 43," Studies in Iconography, xxii (2001), 
pp. 21-53. 

64. Edited by Athanasios Papadopoulos-Kerameus, Manuel d'icono- 
graphie chretienne. (St. Petersburg, 1909). 

65. For suggestions regarding the relationship between picture criti- 
cism and textual criticism, see Kurt Weitzmann, Illustrations in Roll and 
Codex, a Study of the Origin and Method of Text Illustration (Princeton, 
1947), pp. 182-92. 

The Making of Ancient Books 45 

9. Cola and Commata 

The practice of writing treatises in short lines according to the 
sense antedated its application to Christian writings. Some of the ora- 
tions of Demosthenes and Cicero were transcribed in this manner to 
assist the reader in making the correct inflection and the proper 
pauses. It was also applied to the Septuagint Greek text of the poetic 
books of the Old Testament, 66 and when Jerome translated the 
Prophets into Latin, he arranged the text colometrically. 67 Each 
sense-line consisted of a single clause (udoXov) or a single phrase 

A wide variety of bilingual manuscripts of the New Testament 
have survived, 69 the most numerous being manuscripts that present 

66. One of the earliest examples of a portion of the Septuagint 
arranged in cola is the second- (or third-) century Bodleian fragment of 
the Psalms edited by J. W. B. Barns and G. D. Kilpatrick (Proceedings of 
the British Academy, xliii [1957], pp. 229-32). 

67. Jerome, Preface to Isaiah (Migne, Patrologia Latina, xxviii, 
p. 825). 

68. In antiquity, rhetoricians defined a comma as any combination of 
words that makes a total of not more than eight syllables, while they re- 
quired from a colon a combination of at least nine, though not exceeding 
16; see James A. Kleist, "Colometry and the New Testament," Classical 
Bulletin, iv (1928), pp. 26 f: 

these figures will not seem arbitrary if we bear in mind that the average length 
of a hexameter is sixteen syllables, and that the average speaker does not eas- 
ily go beyond sixteen syllables without renewing his breath. ... In writing 
colometrically. the one great purpose agreed upon by all ancient rhetoricians 
is to enable the reader to read intelligently, and properly to husband one's 
breath in speaking. The all-essential thing is that both colon and comma, 
when taken by themselves, make sense and admit of proper breathing. But 
essential as this requirement is, it is also a source of arbitrariness of interpreta- 
tion. How much, after all, or how little is required to make sense? How much 
can be uttered in one breath? Here, as elsewhere, quot capita, tot sententiae, 
and unusquisque in suo sensu abundat. 

See Figure 11, p. 75, for an example of unusually short lines. 

69. See B. M. Metzger, "Bilingualism and Polylingualism in Antiquity; 
with a Check-List of New Testament MSS Written in More than One Lan- 
guage," in The New Testament Age: Essays in Honor of Bo Reicke, ed. by 
William C. Weinrich (Macon, GA, 1984), pp. 327-34. 

46 The Text or the New Testament 

the text in Greek and Latin. 70 Most of these set forth the text in two 
columns, using sense-lines to help the reader correlate the two texts. 
Among other modes of presentation was to write one column evenly 
and the other in lines of corresponding content but uneven length. 
Yet another was to write fewer lines in one column than the other, 
without exact correspondence between lines. A fourth was to pro- 
vide an interlinear translation. 

10. Neumes 

Neumes 71 are Byzantine musical notes that assisted the lector in 
chanting or cantillating the Scripture lesson. They appear first in 
codices of the seventh or eighth century, but whether they were con- 
temporary with the text of the manuscript or added later is difficult 
to determine. Their form is that of hooks, dots, and oblique strokes 
(see Fig. 25, p. 324), and they are usually written with red (or green) 
ink above the words to be sung. 72 

1 1 . Lectionary Equipment 

Following the custom of the synagogue, according to which por- 
tions of the Law and the Prophets were read at divine service each 
Sabbath day, the Christian Church adopted the practice of reading pas- 
sages from the New Testament books at services of worship. A regu- 
lar system of lessons from the Gospels and Epistles was developed, 
and the custom arose of arranging these according to a fixed order of 
Sundays and other holy days of the Christian year. In order to assist the 
lector in finding the beginning (agxrj) and the end (reXog) of the les- 
son, several of the later majuscule manuscripts were provided with the 
abbreviations olq% and reA, inserted either in the margin or between 
the lines of the text (see Fig. 6, p. 37). Lection notes, indicating that a 
given passage is to be read on a certain day, were sometimes written 

70. For a list of 94 nonbiblical Greco-Latin manuscripts and a list of 
25 Greco-Latin New Testament manuscripts, see David C. Parker, Codex 
Bezae: An Early Christian Manuscript and Its Text (Cambridge, 1992), 
pp. 52-4, 60. 

71. See especially H.J. W. Tillyard, Handbook of the Middle Byzantine 
Musical Notation {Monumenta Musicae Byzantinae, Subsidia, 1 ,i; 
Copenhagen, 1935); Carsten Hoeg, La Notation ekphonetique (Monumenta 
Musicae Byzantinae, Subsidia, l.ii; Copenhagen, 1935); and E. G. Wellesz, 
A History of Byzantine Music and Hymnography, 2nd ed. (Oxford, 1961). 

72. See Ewald Jammers, Tafeln zur Neumenschrift, mit einer 
Einfiihrung (Tutzing, 1965). 

The Making of Ancient Books 47 

in the margin with red ink. A list of all these passages is occasionally 
given at the close (or at the beginning) of a codex. 

For the added convenience of the reader, lectionary manuscripts 
were prepared that present in proper sequence (beginning with 
Easter) the text of the several passages of Scripture appointed to be 
read on Sundays, Saturdays, and in some cases weekdays through- 
out the year. Such lectionaries are called synaxaria (see Fig. 5, 
p. 35). Another service book is the menologion, which supplies 
Scripture lessons for feast days, saints' days, and the like, starting 
with the first of September, the beginning of the civil year. Substan- 
tially the same choice of Scripture passages in lectionary manuscripts 
dating from the seventh or eighth century is still followed by the 
Greek Orthodox Church today. 

Scholars have recognized the importance of lectionary manu- 
scripts in tracing the history of the text of the New Testament during 
the Byzantine period. 73 Inasmuch as the form of the citation of the 
Scriptures in official liturgical books always tends to be conservative 
and almost archaic, 74 lectionaries are valuable in preserving a type of 
text that is frequently much older than the actual age of the manu- 
script might lead one to suspect. 

V. Statistics of Greek Manuscripts 
of the New Testament 75 

It is customary to classify the Greek manuscripts of the New 
Testament into several categories, partly according to the material 

73. For an introduction to the study of New Testament Greek lec- 
tionaries, see Ernest C. Colwell and Donald W. Riddle, Prolegomena to the 
Study of the lectionary Text of the Gospels (Chicago, 1933), and more re- 
cently Klaus Junack, "Lectionaries," Anchor Bible Dictionary, vol. 4 (New 
York, 1992), pp. 271-3, and Carroll D. Osburn, "Greek Lectionaries of the 
New Testament," in Ihe Text of the New Testament in Contemporary Re- 
search, ed. by Ehrman and Holmes, pp. 61-74. See also n. 77. 

74. Thus, e.g., until 1928 The Anglican Book of 'Common Prayer ^con- 
tinued to have the Psalter and the Canticles from Luke in die translation of 
the Great Bible of 1539, despite repeated proposals to make them conform 
to the King James Version of 1611. 

75. See Kurt Aland, Kurzgefafite Liste der griechischen Handschriften des 
Neuen Testaments, zweite, neubearbeitate und enganzte Auflage. (Berlin, 

1994). An up-to-date list is kept on the website of the Institut fiir neutesta- 
mentliche Textforschung ( 

48 The Text of the New Testament 

from which they are made, partly according to their script, and partly 
according to the use for which they were intended. 

At first, editors of the New Testament used cumbersome titles to 
designate the Greek manuscripts. These titles were usually derived 
from the names of the owners or the libraries possessing the manu- 
scripts. Since no one system was agreed upon by all editors and 
manuscripts would change owners and locations, it was exceedingly 
confusing to compare the evidence in one critical apparatus with 
that in another. The first step toward the standardization of nomen- 
clature was taken by Swiss scholar Johann Jakob Wettstein, who, in 
his handsome two-volume edition of the Greek New Testament 
published at Amsterdam in 1751-2, employed capital letters to des- 
ignate majuscule manuscripts and Arabic numerals to designate mi- 
nuscule manuscripts. The system now in general use was elaborated 
at the end of the nineteenth century by Caspar Rene Gregory, 76 a 
native Philadelphian who, after receiving his theological training at 
Princeton, went to Germany where he became a professor of the 
New Testament at the University of Leipzig in 1889. Building upon 
Wettstein's system, Gregory devised several other categories of the 
materials. Thus, the manuscripts made of papyrus are listed sepa- 
rately from those made of parchment. Each of them is commonly 
referred to by the Gothic or Old English letter p followed by a small 
superior numeral. At the time of this writing (2003), a total of 116 
Greek papyri of the New Testament have been assigned official 
numbers by Gregory and his successors. 

Following Wettstein's system, the majuscule manuscripts that 
have been known for the longest time are commonly designated in 
a critical apparatus by capital letters of the Latin and Greek alphabets 
and by one Hebrew letter (», aleph). Since, however, the number of 
known majuscule manuscripts came to exceed the number of letters 
in the Latin, Greek, and Hebrew alphabets combined, Gregory as- 
signed to each majuscule manuscript an Arabic numeral preceded by 
a zero. Thus far, 310 majuscule manuscripts have been catalogued. 
The minuscule manuscripts now total 2,877. 

A subsidiary class of Greek manuscripts, involving both majus- 
cule and minuscule manuscripts (though the latter by far predomi- 
nate in number), is devoted to lectionaries. As mentioned earlier, 

76. For biographical details of Gregory's ancestry and life, with a pic- 
ture of the mature scholar, see the journal Biblical World, NS 38, no. 5 
(December 1911), pp. 350-4 (the picture is the frontispiece). 

The Making of Ancient Books 49 

these are church reading books containing the text of selections of 
the Scriptures appointed to be read on the several days of the eccle- 
siastical and the civil year, comprising, respectively, the synaxarion 
and the menologion. Although 2,432 lectionaries of the Greek New 
Testament have been catalogued, only comparatively few have been 
critically studied. 77 In the Gregory system of designating manu- 
scripts, lectionaries are indicated by the letter / followed by an Arabic 
numeral. Thus, / alone designates a Gospel lectionary; / a designates 
a lectionary of the Acts and the Epistles; and / +a designates a lec- 
tionary containing lessons from Gospels, Acts, and Epistles. (The 
Greek lectionary contains no lessons from the Book of Revelation.) 
Two other forms of New Testament witness may be described 
with only a few words. Short portions of six New Testament books 
have been preserved on ostraca, or broken pieces of pottery used by 
the poorest people as writing material. Twenty-five of these have 
been catalogued and are sometimes referred to by the Gothic or Old 
English letter followed by a superior numeral. Finally, a curious 
but unimportant source of our knowledge of the Greek text of the 
New Testament consists of a number of talismans, or good-luck 
charms. These amulets range in date from the fourth to the twelfth 
or thirteenth centuries and are made of vellum, papyrus, potsherd, or 
wood. The superstitious use of talismans, so prevalent in the ancient 
world, was scarcely less popular among Christians than among 
pagans — if we may judge from repeated remonstrances against 
them issued by ecclesiastical authorities. 78 Four of those catalogued 

77. See the series of monographs produced at the University of 
Chicago and published under the general title Studies in the Lectionary 
Text of the Greek New Testament, begun by E. C. Colwell and D. W. Riddle. 
They are J. R. Branton, The Common Text of the Greek Lectionary in the 
Lenten Lections (1934); Morgan Ward Redus, The Text of the Major Festivals 
of the Monologion in the Greek Gospel Lectionary (1936); Bruce M. Metzger, 
'Ihe Saturday and Sunday Lessons from Luke in the Greek Gospel Lec- 
tionary (1944); Harry M. Buck, Jr., Thejohannine Lessons in the Greek 
Gospel Lectionary (1958); William D. Bray, The Weekday Lessons from Luke 
in the Greek Gospel Lectionary (1959); Ray Harms, The Matthean Weekday 
Lessons in the Greek Gospel Lectionary (1966). 

78. Thus, in addition to remonstrances by Eusebius and Augustine, 
the Synod of Laodicea issued a separate canon proscribing the manufacture 
and use of amulets: "and those who wear such we command to be cast out 
of the Church." For these and other references, see the annotations on a 

50 The Text of the New Testament 

contain the Lord's Prayer, and five others include scattered verses 
from other parts of the Old and New Testaments. They are some- 
times referred to by the letter % followed by a superior numeral. 

Disregarding for the moment ostraca and talismans, the official 
listing (as of 2003) of the several important categories of Greek New 
Testament manuscripts can be summarized as follows: 

Papyri 116 

Majuscule MSS 310 

Minuscule MSS 2877 

Lectionary MSS 2432 

Total 5735 

We can appreciate how bountiful the attestation is for the New 
Testament if we compare the surviving textual materials of other an- 
cient authors who wrote during the early centuries of Christianity. 
For example, the compendious History of Rome, written in Latin by 
Velleius Paterculus (bom c. 20 B.C.; died after A.D. 30), survived in 
only one incomplete manuscript, discovered in 1515 at the Benedic- 
tine Abbey of Murback in Alsace. The editio princeps was prepared 
by the humanist Beatus Rhenanus and issued by the printer Froben 
at Basel in 1520. As for the manuscript itself, in the following years it 
was lost without a trace. One other copy, however, had been made 
in 1516 by a pupil of Rhenanus, the Baseler humanist Boniface 
Amerbach. This still exists today in the library of the University of 
Basel, where it was discovered in 1834 by the classical philologist 
Johann Kaspar Orelli. The relative merits of the copy and the first 
printed edition are disputed. 79 

The surviving texts of the famous Latin historian Cornelius Taci- 
tus (flor. c. 100 A.D.) reached the age of printing by three tenuous 

papyrus fever amulet edited by Bruce M. Metzger in Papyri in the Prince- 
ton University Collections, iii (Princeton, 1942), pp. 78 f. (An enlarged edi- 
tion of the fever amulet is given in Metzger's Literary and Historical Studies 
[Leiden, 1968], pp. 104—10.) See also S. R. Pickering, "The Significance of 
Non-Continuous New Testament Textual Materials in Papyri," in Studies in 
the Early Text of the Gospels and Acts, ed. by D. G. K. Taylor (Atlanta, 
1999), pp. 121^0. 

79. A useful bibliographical essay, entitled "Etat present des 
travaux sur l'Histoire Romaine de Velleius Paterculus," was written by 
J. Hellegouac'h for Aufstieg und Niedergang der Rdmischen Welt, 11.32. 1 
(Berlin, 1984), pp. 404-36. 

The Making of Ancient Books 51 

threads. Of the fourteen books of his Histories, only four and a half 
survive; of the sixteen books of his Annals, ten survive in full and 
two in part. The text of these extant portions of his two great histor- 
ical works depends entirely on two MSS, one of the ninth century 
and one of the eleventh. His minor works have all descended from 
a codex of the tenth century, but this disappeared after numerous 
fifteenth-century copies had been made. 

In contrast with these figures, the textual critic of the New 
Testament is embarrassed by the wealth of material. 80 Furthermore, 
the work of many ancient authors has been preserved only in man- 
uscripts that date from the Middle Ages (sometimes the late Middle 
Ages), far removed from the time at which they lived and wrote. On 
the contrary, the time between the composition of the books of the 
New Testament and the earliest extant copies is relatively brief. In- 
stead of the lapse of a millennium or more, as is the case of not a few 
classical authors, several papyrus manuscripts of portions of the New 
Testament are extant that were copied within a century or so after 
the composition of the original documents. For further details, see 
the next chapter. 

80. Lest, however, the wrong impression be conveyed from the statis- 
tics given above regarding the total number of Greek manuscripts of the 
New Testament, it should be pointed out that most of the papyri are 
relatively fragmentary and that only about 60 manuscripts (of which Codex 
Sinaiticus is the only majuscule manuscript) contain the entire New 
Testament. The great majority of the other manuscripts contain only the 
four Gospels or only the Epistles. The Book of Revelation is the least well- 
attested part of the New Testament, being preserved in about 300 Greek 
manuscripts. Of this number, only ten are majuscule manuscripts (i.e., «, A, 
C, P, 046, 051, 052, 0163, 0169, and 0207), and three of these ten comprise- 
only a single leaf each (the last three mentioned). 

A significant number of copies of the Book of Revelation (at least 35) 
are included in the same codex with nonbiblical documents; for a list, see 
B. M. Metzger, "The Future of New Testament Studies," in The Bible as 
Book: The Transmission of the Greek Text, ed. by Scott McKendrick and 
Orlaith O'SulUvan (London, 2003), pp. 205-6. 


Important Witnesses to the Text 
of the New Testament 

Three classes of witness are available for ascertaining the text of the 
New Testament; they are the Greek manuscripts, the ancient transla- 
tions into other languages, and the quotations from the New Testa- 
ment made by early ecclesiastical writers. Something must be said 
now about certain individual witnesses in each of these three classes 
of evidence. 

I. Greek Manuscripts of the New Testament 

Of the approximately 5,700 Greek manuscripts that contain all or 
part of the New Testament, the following are among the most im- 
portant. They are listed here under the usual categories of (1) papyri, 
(2) majuscules, and (3) minuscules; within each of these groups, the 
sequence is that of the Gregory system of numeration. In the 
descriptions of these manuscripts, reference is frequently made to 
various types of text, such as the Alexandrian, the Western, the 
Caesarean, and the Koine or Byzantine; for information concerning 
the significance of such terminology, see pp. 306-13- The name of 
the editor or collator is given for manuscripts that were published 
individually; a more or less full conspectus of readings of the other 
manuscripts mentioned here may be found in the standard appa- 
ratus critici. 


Important Witnesses to the Text of the New Testament 53 

1. Important Greek Papyri of the New Testament 1 

P 4 , P 64 , P 67 When first discovered and catalogued, these papyrus 
fragments were not recognized as belonging originally to the same 
manuscript. The careful study by Skeat, however, has shown that 
they all derive from a single-quire codex that originally contained all 
four Gospels, either in the canonical order of Matthew, Mark, Luke, 
and John or the so-called Western order of Matthew, John, Luke, and 
Mark. 2 

p 4 consists of portions of four leaves from the early chapters of 
Luke; p 64 is a fragment of a single leaf containing verses from 
Matthew 26; and p 67 consists of two fragments, the first with por- 
tions of Matt. 3.9 and 3-25 and the other with portions of Matt. 
5.20-2 and 5.25-8. The manuscript was written in two columns with 
36 lines in each. Among its significant features is its use of an orga- 
nized text division, in which a new section of text (such as a new 
paragraph) begins with a colon combined with the projection into 
the left margin of the initial letter or letters of the next complete 
line. Skeat dated the fragments to the "late second century." As 
such, they represent the oldest four-Gospel manuscript known to 
exist and push the practice of organized text division back into the 
second century. 

Two of the most important collections of papyrus manuscripts of 
the New Testament were acquired by Sir Chester Beatty of London 
in 1930-1 and by Martin Bodmer of Geneva in about 1955-6. The 
former collection is now in the Chester Beatty Library, in a suburb of 
Dublin, and has been edited, with introductions and discussions, by 
Sir Frederic G. Kenyon. 3 

1. A useful volume that provides the Greek text, along with many 
paleographical annotations, of 65 papyrus and four parchment fragments 
dated prior to A.D. 300 is The Text of the Earliest New Testament Greek 
Manuscripts, ed. by Philip W. Comfort and David R Barnet (Wheaton, IL, 
2001). For a helpful overview, see Eldon Jay Epp, "The Papyrus Manu- 
scripts of the New Testament," in The New Testament in Contemporary Re- 
search: Essays on the Status Quaestionis ed. by Bart D. Ehrman and Michael 
W. Holmes (Grand Rapids, MI, 1995), pp. 3-21. 

2. T. C. Skeat, "The Oldest Manuscript of the Four Gospels?" New 
Testament Studies, xliii (1997), pp. 1-34. 

3. The Chester Beatty Biblical Papyri, Descriptions and Texts (London, 

54 The Text of the New Testament 

p 45 . The first of the Chester Beatty biblical papyri, to which the 
siglum p 4? has been assigned, comprises portions of 30 leaves of a 
papyrus book formed by a succession of quires of only two leaves. 
Originally the codex consisted of about 220 leaves, each measuring 
about 10 by 8 inches, and contained all four Gospels and the Acts. 
Today, Matthew and John are the least well preserved, each being 
represented by only two fragmentary leaves. Six leaves of Mark, 
seven of Luke, and 13 of Acts remain of these books. Several small 
fragments of the codex, originally comprising a leaf from Matthew, 
have turned up in a collection of papyri at Vienna. 4 

The manuscript is dated by the editor in the first half of the third 
century. The type of New Testament text that it preserves in Mark is 
nearer to the Caesarean family than to either the Alexandrian or the 
Western text types. In the other Gospels (where the Caesarean text 
has not yet been fully established), it is also intermediate between 
the Alexandrian and the Western. In Acts, it is decidedly nearer the 
Alexandrian and has none of the major variants characteristic of the 
Western text of this book, though it has some of the minor ones. 

p 46 . The second Chester Beatty biblical papyrus, designated p 46 , 
comprises 86 leaves (all slightly mutilated) of a single-quire 5 pa- 
pyrus codex, measuring originally about 11 by 6'A inches, which 
contained on 104 leaves ten Epistles of Paul in the following order: 
Romans, Hebrews, 1 and 2 Corinthians, Ephesians, Galatians, Philip- 
pians, Colossians, 1 and 2 Thessalonians. Slightly earlier than p 45 , 
it dates from about A.D. 200. Today, portions of Romans and 1 Thes- 
salonians as well as 2 Thessalonians in its entirety are lacking. The 
Pastoral Epistles were probably never included in the codex, for there 
does not appear to be room for them on the leaves missing at the end. 

4. Edited by Hans Gerstinger, "Ein Fragment des Chester Beatty- 
Evangelienkodex in der Papyrussammlung der Nationalbibliothek in 
Wien," Aegyptus, xiii (1933), pp. 67-72. For further codicological studies of 
the Vienna fragments, see T. C. Skeat and B. C. McGing in llermathena, cl 
(199D, pp. 21-5; Skeat, Hermathena, civ (1993), pp. 27-43; T.J. Kraus, 
Biblica, lxxxii (2001), pp. 1-15. 

5. Three (possibly four) of the Beatty biblical papyri are single-quire 
manuscripts. For a discussion of this form of codex, see Campbell Bonner's 
introduction to his Papyrus Codex of the Shepherd of Hermas (Ann Arbor, 
MI, 1934), pp. 7-12. More recently, several other examples of single-quire 
codices have come to light, including the Michigan Gospel of John in 
Fayyumic Coptic and 12 of the 13 Gnostic manuscripts from Nag Hammadi. 

Important Witnesses to the Text of the New Testament 55 

(Since it is a single-quire codex, the number of leaves lacking at both 
ends can be computed more or less accurately.) Thirty of the 86 sur- 
viving leaves are in the library at the University of Michigan.' 

In addition to the reversal of the present order of Galatians and 
Ephesians, the anonymous Epistle to the Hebrews is included among 
the Pauline Epistles, which are arranged in general order of decreas- 
ing length. p 4fi is noteworthy, likewise, in that the doxology to 
Romans, which in many of the earlier manuscripts stands at the end 
of chapter 14, is placed at the end of chapter 15 (see Fig. 7, p. 42). 7 
In general, the papyrus is closer to the Alexandrian than to the West- 
ern type of text. 

p 47 . The third Chester Beatty biblical papyrus of the New Testa- 
ment, designated p 47 , comprises ten slightly mutilated leaves of a 
codex, measuring about 9'A by 5'A inches, of the Hook of Revelation. 
Of the original codex, estimated to have been 32 leaves in length, only 
the middle portion remains, containing the text of 9-10-17.2. It dates 
from the middle or latter part of the third century. In general, the text 
of p 47 agrees more often with that of Codex Sinaiticus than with any 
other, though it often shows a remarkable independence. 

p 52 . Measuring only 2'A by y/i inches and containing but a few 
verses from the fourth Gospel (18.31-3, 37-8), this papyrus fragment 
is the oldest copy of any portion of the New Testament known to be 
in existence today. Although it was acquired in Egypt by Bernard P. 
Grenfell as long ago as 1920, it remained unnoticed among hundreds 
of similar shreds of papyri until 1934. In that year, C. H. Roberts, 
Fellow of St. John's College, Oxford, while sorting over the unpub- 
lished papyri belonging to the John Rylands Library at Manchester, 
recognized that this scrap preserves several sentences from John. 

6. Edited by Henry A. Sanders, A Third-Century Papyrus Codex of the 
Epistles of Paul (Ann Arbor, Ml, 1935). Sir Frederic G. Kenyon re-edited the 
30 leaves of the portion in the possession of the University of Michigan, 
along with ten leaves that he had previously edited, to which were added 
46 newly acquired leaves of the same codex, in Fasciculus iii Supplement, 
Pauline Epistles, of The Chester Beatty Biblical Papyri (London, 1936). 

7. This papyrus, however, contrary to common opinion, is not alone 
in placing Hebrews immediately following Romans; in six minuscule man- 
uscripts and in a Syrian canon composed about A.D. 400, Hebrews occu- 
pies this position; see W. H. P. I latch, "The Position of I Iebrews in the 
Canon of the New Testament," Harvard 'theological Review, xxix (1936), 
pp. 133-51. 

56 The Text of the New Testament 

Without waiting to edit the fragment along with others of a miscella- 
neous nature, he immediately published a booklet setting forth a 
description of the fragment, its text, and its significance. 8 

On the basis of the style of the script, Roberts dated the fragment 
to the first half of the second century. Though not all scholars are 
convinced that it can be dated within so narrow a range, such emi- 
nent paleographers as Sir Frederic G. Kenyon, W. Schubart, Sir 
Harold I. Bell, Adolf Deissmann, Ulrich Wilcken, and W. H. P. Hatch 
have expressed agreement with Roberts' judgment. 9 

Although the extent of the verses preserved is so slight, in one 
respect this tiny scrap of papyrus possesses quite as much evidential 
value as would the complete codex. Just as Robinson Crusoe, seeing 
but a single footprint in the sand, concluded that another human 
being, with two feet, was present on the island with him, so \P 2 
proves the existence and use of the fourth Gospel during the first 
half of the second century in a provincial town along the Nile, far re- 
moved from its traditional place of composition (Ephesus in Asia 
Minor). Had this little fragment been known during the middle of the 
past century, that school of New Testament criticism which was in- 
spired by the brilliant Tubingen professor Ferdinand Christian Baur 
could not have argued that the fourth Gospel was not composed 
until about the year 160. 

p 66 . The most important discoveries of New Testament manu- 
scripts since the purchase of the Chester Beatty papyri are the ac- 
quisitions made by the Genevan bibliophile and humanist Martin 
Bodmer, founder of the Bodmer Library of World Literature at 
Cologny, a suburb of Geneva. One of the oldest considerable por- 
tions of the Greek New Testament is a papyrus codex of the Gospel 
of John, Bodmer Papyrus II, which was published in 1956 by Victor 
Martin, professor of classical philology at the University of Geneva. 

8. C. H. Roberts, An Unpublished Fragment of the Fourth Gospel in the 
John Rylands Library (Manchester, 1935). This was republished, with slight 
alterations, in the Bulletin of the John Rylands Library, xx (1936), 

pp. 45-55, and again, with critical notes and a bibliography of reviews and 
opinions expressed by other scholars, in the Catalogue of the Greek and 
Latin Papyri in the John Rylands Library, iii (Manchester, 1938), pp. 1-3. 

9. Deissmann was convinced that p 52 was written well within the reign 
of Hadrian (A.D. 117-38) and perhaps even during the time of Trajan (A.D. 
98-117); see his article "Ein Evangelienblatt aus den Tagen Hadrians," 
Deutsche allgemeine Zeitung, 564 (3 Dec. 1935); English trans, in British 
Weekly (12 Dec. 1935), p. 219. 

Important Witnesses to the Text of the New Testament 57 

According to its editor, the manuscript dates from about A.D. 200. 10 
It measures about 6 by 5^ inches and consists of six quires, of which 
104 pages remain. These contain the text of John 1.1-6.11 and 
6.35b— 14.15. Subsequently, fragments of 46 other pages of the same 
codex were also acquired by Bodmer and edited as a supplement by 
Martin (1958). n Since most of these fragments are small, some of 
them mere scraps, the amount of text of John 14-21 that has been 
preserved is not great. 

The text of p 66 is mixed, with elements that are typically Alexan- 
drian and Western. The manuscript contains about 440 alterations, 
introduced between lines, over erasures, and in the margins. Most 
of them appear to be the scribe's corrections of his own hasty blun- 
ders, though others seem to imply the use of a different exemplar. 
Several passages present unique readings that previously had not 
been found in any other manuscript. In 13.5, a picturesque word is 
used in connection with the washing of the disciples' feet; accord- 
ing to p 66 , Jesus took not a "basin" (vuiTfJQa) but a "foot-basin" 
(nodoviTZTiJQa). In 7.52 the presence of the definite article in a diffi- 
cult passage now supports what scholars had long thought was the 
required sense, namely "Search [the Scriptures] and you will see that 
the prophet does not rise from Galilee." 12 

10. Herbert Hunger, the director of the papyrological collections in the 
Austrian National Library at Vienna, dated p 66 earlier, in the middle if not 
even in the first half of the second century; see his article "Zur Datierung 
des Papyrus Bodmer II (p 66 )," Anzeiger der osterreichischen Akademie der 
Wissenschaften, phil.-hist. Kl., 4 (I960), pp. 12-33- 

11. A new edition of the supplement, augmented and corrected, was 
published in 1962 with the assistance of J. W. B. Barns of Oxford, accom- 
panied by a photographic reproduction of the entire manuscript (chapters 
i-xxi). For still further emendations, see Barns, "Papyrus Bodmer II, Some 
Corrections and Remarks," Museon, lxxv (1962), pp. 327-9; E. F. Rhodes, 
"The Corrections of Papyrus Bodmer II," New Testament Studies, xiv 
(1967-8), pp. 271-81. 

12. For further studies of p 66 , see J. N. Birdsall, The Bodmer Papyrus of 
the Gospel of John (London, I960); M.-E. Boismard in Revue Biblique, lxx 
(1963), pp. 120-37; G. D. Fee, "The Corrections of Papyrus Bodmer II and 
Early Textual Transmission," Novum Testamentum, vii (1965), pp. 247-57; 
idem, Papyrus Bodmer 11 (p 66 ): Its Textual Relationships and Scribal Charac- 
teristics {Studies and Documents 34; Salt Lake City, UT, 1968); Miguel Balgue 
in Studia papyrologica, iv (1965), pp. 76-89; E. C. Colwell, "Scribal Habits in 
Early Papyri; A Study in the Corruption of the Text," in The Bible in Modern 
Scholarship, ed. by J. Philip Hyatt (Nashville, TN, 1965), pp. 370-89- 

58 The Text or the New Testament 

p 72 . The earliest known copy of the Epistle of Jude and the two 
Epistles of Peter is contained in another papyrus codex that was ac- 
quired by Bodmer and edited by Michel Testuz (1959). This manu- 
script, which the editor dates to the third century, contains a miscel- 
laneous assortment of documents in the following order: the Nativity 
of Mary, the apocryphal correspondence of Paul to the Corinthians, 
the eleventh Ode of Solomon, the Epistle of Jude, Melito's homily on 
the Passover, a fragment of a hymn, the Apology of Phileas, Psalms 
33 and 34, and the two Epistles of Peter. Because of the relatively 
small size of the codex (6 by 5% inches), the editor concluded that it 
was made for private usage and not for reading in church services. 
Apparently, four scribes took part in producing the manuscript. The 
affinities of its text of 1 Peter belong definitely with the Alexandrian 
group, particularly with the Codex Alexandrinus. 

p 74 . Bodmer Papyrus XVII, edited by Rodolphe Kasser in 1961, is a 
rather voluminous codex dating from the seventh century. Originally, 
the manuscript contained 264 pages, each measuring about 13 by 
8 inches. Today, it is in a poor state of preservation and contains, with 
many lacunae, portions of the Book of Acts; James; 1 and 2 Peter; 1, 2, 
and 3 John; and Jude. The type of text that it preserves agrees fre- 
quently with Alexandrian witnesses. 13 

p 75 . Still another early biblical manuscript acquired by Bodmer is 
a single-quire codex of Luke and John. It originally contained about 
144 pages, each measuring 10% by 5J<* inches, of which 102 have sur- 
vived, either in whole or in part. The script is a clear and carefully 
executed majuscule, somewhat like that of p 4, 5 though with a less 
pronounced ductus. The editors, Victor Martin and Rodolphe Kasser, 
date this copy between A.D. 175 and 225. It is thus the earliest 
known copy of the Gospel according to Luke and one of the earliest 
of the Gospel according to John. The orthography of the name John 
varies in a curious manner. In Luke, it is invariably written with a sin- 
gle v Claxivrjg), and the same orthography appears at the beginning 
of the Gospel according to John. At John 1.26, however, a second v 
is added between the lines above the a and the v (as also at 10.40), 
and thereafter the geminated form appears everywhere except at 

13- According to Philippe-H. Menoud, in Acts p 74 agrees more fre- 
quently with S and A than with B, especially as to order of words; it sup- 
ports no truly Western reading ("Papyrus Bodmer XVII," Revue de theologie 
et de philosophie, 3rd ser., xii [1962], pp. 112-16). 

Important Witnesses to the Text of the New Testament 59 

3.27, where, perhaps because of a distraction, the scribe reverted to 
his former spelling. 

The textual significance of this witness is hard to overestimate, 
presenting as it does a form of text very similar to that of Vaticanus. 14 
Occasionally, it is the only known Greek witness that agrees with the 
Sahidic in supporting several interesting readings. Thus, at John 10.7, 
instead of the traditional text "I am the door of the sheep," p 75 re- 
places "door" (77 Ovga) with "shepherd" (0 jzoiftJjv). What is still 
more remarkable is the addition at Luke 16.19, where in Jesus' ac- 
count of the rich man and Lazarus this witness inserts after nkovaiog 
the words ovopiari Nevtjg (see Fig. 8). The Sahidic version agrees 
with a rather widespread tradition among ancient catechists of the 
Coptic Church that the name of the rich man was "Nineveh," a name 
that had become the symbol of dissolute riches. Obviously, the 
scribe of p 75 was acquainted with this tradition and by accidental 
haplography wrote "Neve" for "Nineveh" (Nevrjc, for Nivevrjc,). 15 

14. For further studies of p 75 , see C. L. Porter, "Papyrus Bodmer XV 
(p 7 " 1 ) and the Text of Codex Vaticanus," Journal of Biblical Literature, 
lxxxi (1962), pp. 363-76; Kurt Aland, "Neue neutestamentliche Papyri," 
New Testament Studies, xi (1964-5), pp. 5-21; idem, New Testament 
Studies, xii (1965-6), pp. 195-210 (reprinted, with additions, in Aland's 
Studien zur Uberlieferung des Neuen Testaments und seines Textes 
[Berlin, 19671, pp. 155-72); Gordon D. Fee, "p 75 , p 66 , and Origen: The 
Myth of Early Textual Recension in Alexandria," in New Dimensions in 
New Testament Study ed. by R. N. Longenecker and M. C. Tenney 
(Grand Rapids, MI, 1974), pp. 17-45; and especially Carlo M. Martini, // 
problema della recensionalita del codice B alia luce del papiro Bodmer 
A7K(Rome, 1966). 

15. It was probably the horror vacui that led more than one reader to 
provide a name for the anonymous rich man. Toward the close of the 
fourth century Priscillian, a highly educated layman who revived certain 
Manichean errors in southern Spain, gave the name "Finees" to the rich 
man, perhaps because in the Old Testament Eleazar (compare "Lazarus") 
and Phinehas are associated. (The only manuscript extant of Prise. Tract, xi 
[ed. by G. Schepps, p. 91] reads "Fineet," with the t cancelled and an 5 
written above.) 

The widespread use of "Dives" as the name of the rich man is, of 
course, to be accounted for by the rendering of the passage in the Latin 
Vulgate: "Homo quidam erat dives et . . ." Among the literature on the 
subject, see A. Harnack, "Der Name des reichen Mannes in Luc 16, 19," 
Texte und Untersuchungen, xiii (1) (1895), pp. 75-8; J. Rendel Harris, 

«•* • •■■ J» • 

cr* f»K«y« iHrrsyKOJ^xtuHJcrH*!^ 

VI TpUi-.C H C~T»X AJ <*> Nil .yt rKWvlJkC" * 
,rnO"»'frW*J'JvkjcS-«wKIJf*' r " T *-^- f ■ 

A^J h> c * VJ Wg tt • AAfOAA j /<:• e *0» w B 

jjj ki nicrr*-yp*}">cH h* xrjwiXVn-j/ 
oo i-r > *~r« i »>-K t-rtKi re •f*r*y*Jlf*T*f f 
— j-JVyXJUNi-ioo ipJ «T-^*J c»)K»-irH «>y- 
Aj*rivi.xyeiKy/i.;cA»y^Y*J > >> > "' 

rXf-^w^,MJl^rMtHKArT»Wt- ; ~r« i T- 

/■) c4J*| J,)A*J»r)-^»f> i rWLr->f ,; '' r T c,a ' 

Ay^icrUHr trrf-JAikVJ-y^-T^ 
tX.y-n.yc w»»i-t I- w-T^^'*7 pJ j. c 

\H &\FA)f^m^>g<*, rrl -xrT»r*T'« 

» rv r»A) ^ *~txj KTxn-rJitH ay-** 1 **^- 

r-f"*J -yp^M • K»K*r-rW#sirf)H rt*J»»>H ' 

TTJ^ .Xtt .Ay '-Kr-rM N/V>Al *vA ^TSV; 
Wj »"AJU*»" p-T^/OxM.!. Urf-y »» KAI A^j 
rr«Aft\)-AJ>»JM»iJLnri An Af »» >-*Jii"»*> 
JU»'-V«y*»'- XNtfftorTteJ^TKW^ 

f J*|A)TJ»riNp* MH^Aji'AiwafrrKii'!/; «' 

4.' • • ■ *.' ■ '. 41 

l'I te ore 8 Bodmer Papyrus XIV (p 75 , about A.D. 175-225), Cologny/Geneva. 
Luke 16.9-21 (the name of the rich man is given in line 8 from the bottom; 
see p. 59). Actual size 10% X 5% inches. 


Important Witnesses to the Text of the New Testament 6l 

p 115 . This is a fragmentary text of the Book of Revelation, re- 
cently published as Oxyrhynchus papyrus 4499. 16 It consists of 
26 fragments from nine different pages; it is impossible to know 
whether the original manuscript included other texts along with Rev- 
elation. The fragments can be paleographically dated to the late third 
or early fourth century, making this one of the oldest witnesses to 
the Book of Revelation, somewhat older than Codex Sinaiticus but 
not as old as p 47 . More important, the fragments evidence a high 
quality of text. They frequently align with manuscripts A and C, mak- 
ing p m "the oldest member by over a century of the A C textype"; 17 
they often support A in variant readings that can be judged to repre- 
sent the oldest form of the text. 

Among its many interesting features is that p 115 is partially extant 
at Rev. 13.18, a passage in which, according to most manuscripts, the 
number of the anti-Christ is given as 666. In p 115 , however, along 
with manuscript C and witnesses known to Irenaeus, the number is 
616. (It is interesting to note that if "Caesar Neron" is spelled in 
Hebrew letters, their numerical value is 666 — unless, that is, the op- 
tional nun is omitted at the end, in which case the total is 6l6.) 
Parker argues that, in a number of other instances, readings of this 
fragmentary papyrus should affect the selection of the text of Reve- 
lation in the printed editions. 

"On Certain Obscure Names in the New Testament," Expositor, 6th ser., i 
(1900), pp. 161-77, 304-8, especially pp. 175-7; Hugo Gressmann, Vom 
reichen Mann und armen Lazarus: eine literargeschichtliche Studie 
(Abbandlungen der preussischen Akademie der Wissensch often, phil.-hist. 
Kl., 1918): A. Meyer, "Namen der Namenlosen," in E. Hennecke, Neutesta- 
mentliche Apokryphen, 2te Aufl. (Tubingen, 1924), pp. 78-81; L. Th. Lefort, 
"Le Nom du mauvais riche (Lc 16. 19) et la tradition copte," Zeitschrift fur 
die neutestamentliche Wissenschaft, xxxvii (1938), pp. 65-72; Henry J. 
Cadbury, "A Proper Name for Dives," Journal of Biblical Literature, lxxxi 
(1962), pp. 399^02; Bruce M. Metzger, "Names for the Nameless: A Study 
in the Growth of Christian Tradition," in Kyriakon: festschrift Johannes 
Quasten, ed. by Patrick Granfleld and Josef A. Jungmann (Munster, 1970), 
pp. 79-99. 

16. N. Gonis, et al. The Oxyrhynchus Papyri, lxvi (London, 1999), 
pp. 10-35 and Plates III— VIII, XI-XII. See also David C. Parker, "A New 
Oxyrhynchus Papyrus of Revelation: p 115 (P.Oxy. 4499)," New Testament 
Studies, xlvi (2000), pp. 159-74. 

17. Parker, op. cit., p. 174. 

62 The Text of the New Testament 

2. Important Greek Majuscule Manuscripts 
of the New Testament 

8. Primacy of position in the list of New Testament manuscripts 
is customarily given to the fourth-century codex of the Greek Bible 
discovered about the middle of the nineteenth century by Dr. 
Constantin von Tischendorf at the monastery of St. Catherine on 
Mount Sinai. Hence, this manuscript is known as "Codex Sinaiticus." 
It once contained the entire Bible written in a carefully executed ma- 
juscule hand (see Fig. 2, p. 19) and arranged with four columns per 
page, measuring about 15 by 13^ inches. Today, parts of the Old 
Testament have perished, but fortunately the entire New Testament 
has survived. In fact, Codex Sinaiticus is the only known complete 
copy of the Greek New Testament in majuscule script. 

The story of its discovery is fascinating and deserves to be told 
in some detail. In 1844, when he was not yet 30 years of age, 
Tischendorf, a Privatdozent at the University of Leipzig, began an 
extensive journey through the Near East in search of biblical manu- 
scripts. While visiting the monastery of St. Catherine at Mount Sinai, 
he chanced to see some leaves of parchment in a wastebasket full of 
papers destined to ignite the fire for the oven. On examination, these 
proved to be pan of a copy of the Septuagint version of the Old 
Testament, written in an early Greek majuscule script. He retrieved 
from the basket no fewer than 43 such leaves, and the monk casu- 
ally remarked that two basketloads of similarly discarded leaves had 
already been burned up! Later, when Tischendorf was shown other 
portions of the same codex (containing all of Isaiah and 1 and 4 Mac- 
cabees), he warned the monks that such papers were too valuable to 
be used to stoke their fires. The 43 leaves that he was permitted to 
keep contained portions of 1 Chronicles, Jeremiah, Nehemiah, and 
Esther; and upon returning to Europe, he deposited them in the 
university library at Leipzig, where they still remain. In 1846, he 
published their contents, naming them the "Codex Frederico- 
Augustanus" (in honor of the king of Saxony, Frederick Augustus, 
the discoverer's sovereign and patron). 

In 1853, Tischendorf revisited the monastery of St. Catherine, 
hoping to acquire other portions of the same manuscript. The ex- 
citement that he had displayed on the occasion of his discovery dur- 
ing his first visit had made the monks cautious, and he could learn 
nothing further about the manuscript. In 1859, his travels took him 
back once more to Mount Sinai, this time under the patronage of the 

Important Witnesses to the Text of the New Testament 63 

czar of Russia, Alexander II. The day before he was scheduled to 
leave, he presented to the steward of the monastery a copy of the 
edition of the Septuagint that he had recently published in Leipzig. 
Thereupon, the steward remarked that he too had a copy of the 
Septuagint and produced from a closet in his cell a manuscript 
wrapped in red cloth. There, before the astonished scholar's eyes, 
lay the treasure that he had been longing to see. Concealing his feel- 
ings, Tischendorf casually asked permission to look at it further that 
evening. Permission was granted, and upon retiring to his room 
Tischendorf stayed up all night studying the manuscript — for, as he 
declared in his diary (which, as a scholar, he kept in Latin), quippe 
dormire nefas videbatur ("it really seemed a sacrilege to sleep"). He 
soon found that the document contained much more than he had 
even hoped, for not only was most of the Old Testament there but 
also the New Testament was intact and in excellent condition, with 
the addition of two early Christian works of the second century, the 
Epistle of Barnabas (previously known only through a very poor 
Latin translation) and a large portion of the Shepherd of Hermas, 
hitherto known only by title. 

The next morning, Tischendorf tried to buy the manuscript but 
without success. Then, he asked to be allowed to take it to Cairo to 
study; but the monk in charge of the altar plate objected, so he had 
to leave without it. 

Later, while in Cairo, where the monks of Sinai also had a small 
monastery, Tischendorf importuned the abbot of the monastery of 
St. Catherine, who happened to be in Cairo at the time, to send for 
the document. Thereupon, swift Bedouin messengers were sent to 
fetch the manuscript to Cairo, and it was agreed that Tischendorf 
would be allowed to have it quire by quire (i.e., eight leaves at a 
time) to copy it. Two Germans who happened to be in Cairo and 
who knew some Greek, an apothecary and a bookseller, helped him 
transcribe the manuscript; and Tischendorf revised carefully what 
they copied. In 2 months, they transcribed 110,000 lines of text. 

The next stage of the negotiations involved what may be called 
euphemistically "ecclesiastical diplomacy." At that time, the highest 
place of authority among the monks of Sinai was vacant. Tischendorf 
suggested that it would be to their advantage if they made a gift to 
the czar of Russia, whose influence, as protector of the Greek 
Church, they desired in connection with the election of the new 
abbot — and what could be more appropriate as a gift than this 
ancient Greek manuscript! After prolonged negotiations, the precious 

64 The Text of the New Testament 

codex was delivered to Tischendorf for publication at Leipzig and for 
presentation to the czar in the name of the monks. In the East, a gift 
demands a return (see Genesis 23, where Ephron "gives" Abraham 
a field for a burying plot but nevertheless Abraham pays him 
400 shekels of silver for it). In return for the manuscript, the czar pre- 
sented to the monastery a silver shrine for St. Catherine, a gift of 7,000 
rubles for the library at Sinai, a gift of 2,000 rubles for the monks in 
Cairo, and several Russian decorations (similar to honorary degrees) 
for the authorities of the monastery. In 1862, on the thousandth an- 
niversary of the founding of the Russian Empire, the text of the man- 
uscript was published in magnificent style at the expense of the czar 
in four folio volumes, being printed at Leipzig with type cast for the 
purpose so as to resemble the characters of the manuscript, which it 
represents line for line with the greatest attainable accuracy. I8 

The definitive publication of the codex was made in the twenti- 
eth century, when Oxford University Press issued a facsimile from 
photographs taken by Professor Kirsopp Lake (New Testament, 
1911; Old Testament, 1922). After the revolution in Russia, the 
U.S.S.R., not being interested in the Bible and being in need of 
money, negotiated with the trustees of the British Museum for the 
sale of the codex for £100,000 (then slightly more than $500,000). 
The British government guaranteed one-half the sum, while the 

18. Certain aspects of the negotiations leading to the transfer of the 
codex to the czar's possession are open to an interpretation that reflects 
adversely upon Tischendorf s candor and good faith with the monks at 
St. Catherine's. For an account intended to exculpate him of blame, see 
Erhard Lauch, "Nichts gegen Tischendorf," Bekenntnis zur Kirche: Festgabe 
fur Ernst Sommerlath zum 70. Geburtstag (Berlin, c. 1961), pp. 15-24; for 
an account that includes a hitherto unknown receipt given by Tischendorf 
to the authorities at the monastery promising to return the manuscript from 
St. Petersburg "to the Holy Confraternity of Sinai at its earliest request," see 
Ihor Sevcenko's article "New Documents on Tischendorf and the Codex 
Sinaiticus," Scriptorium, xviii (1964), pp. 55-80. For other assessments, see 
K. Aland, "Konstantin von Tischendorf (1815-1874). Neutestamentliche 
Textforschung damals und heute," Sitzungsberichte der sachsischen 
Akademie der Wissenschaften, phil.-hist. Kl., Bd. 133, Heft 2 (Leipzig, 
1953); J. K. Elliott, Codex Sinaiticus and the Simonides Affair (Thessaloniki, 
1982); James Bentley, Secrets of Mount Sinai; The Story of the World's 
Oldest Bible — Codex Sinaiticus (Garden City, NY, 1986); Archimandrite 
Augustin, "'Codex Sinaiticus' of the Bible; the History of Its Discovery," 
Journal of the Moscow Patriarchate, 1988 (1), pp. 63-8. 

Important Witnesses to the Text of the New Testament 65 

other half was raised by popular subscription, contributions being 
made by interested Americans as well as individuals and congrega- 
tions throughout Britain. Just before Christmas Day, 1933, the manu- 
script was carried under guard into the British Museum. 19 A most 
thorough paleographical study of the manuscript was then under- 
taken by H. J. M. Milne and T. C. Skeat of the museum staff, and their 
results were published in a volume entitled Scribes and Correctors of 
Codex Sinaiticus (London, 1938). 20 Additional information regarding 
the manuscript was brought to light. For example, the application of 
a new technique of manuscript study, the use of ultraviolet-ray 
lamps, enabled Milne and Skeat to discover that when the original 
scribe finished writing John 21.24 he drew two decorative lines (a 
coronis) at the lower part of the column of writing and then ap- 
pended a subscription signifying that the text of the Gospel of John 
was finished. (Similar decorative lines and subscriptions appear else- 
where in the manuscript at the end of books.) Later, the same scribe 
washed the vellum and added the concluding verse (verse 25), 
repeating the coronis and subscription at a correspondingly lower 
position (see Fig. 9). 

What Skeat called the last chapter in the romantic story of 
Codex Sinaiticus 21 involves the discovery in 1975 of a sealed room 
in the monastery of St. Catherine that contained art treasures and 
more than 1,000 manuscripts in various languages, of which 836 
were written in Greek. Among the latter were 12 complete leaves 
(together with some fragments) from Codex Sinaiticus. After making 
a preliminary survey of the extent of the newly found material, 
Professor Panayotis Nikolopoulos, keeper of manuscripts at the 
National Library at Athens, along with conservation experts began 
organizing other scholars in Greece for the vast amount of work 
that would be necessary. 22 

19. The British Library, where the manuscript is now displayed, was 
separated from the British Museum in 1973. 

20. See also Christian Tindall, Contributions to the Statistical Study of 
the Codex Sinaiticus (Edinburgh, 1961), and A. Q. Morton, "Codex Sinaiticus 
Revisited," Irish Biblical Studies, xxiv (2002), pp. 14—31. 

21. T. C. Skeat, "The Last Chapter in the History of Codex Sinaiticus," 
Novum Testamentum, xlii (2000), pp. 313-15, based inter alia on Linos 
Politis, "Nouveaux manuscripts grecs decouverts au Mount Sinai," Scripto- 
rium, xxxiv (1980), pp. 5-17. 

22. For a popular account, see Bentley, op. cit., pp. 187-202. 


The Text of the New Testament 

( m.MAl(»A<H)M( 
Wt'fA iMMtH'KX) 



A ► 1 1 lY-KAl AAAOfh, 


Ml « (A-l lt>l< ( IMUJ 
>i< IflM^! IAJXUA 
Aliyt INAfK>Y<IN 
Ay IU*.'J"Xt»M< OAK 
KMCANfc'lt.' ItH IA-I 
iNKKIiNi h'INlllll 
NYK'llt i-uiiiXW £*• 
• MMni irU>|Al A-nkn* 
I 'MOM 4 Nl HiviiiF 

1 I t I lONfl I I>.KIN- 

o^ MCN'nnn nu> 

iMli'lMMHI 'Al- 

Tiftt l'iiN' 

Acri'ioyN AY'MHCr 
1 lAlAlXM»"ift>t 

vl-A| lOI I* '.WH*!!' 

Vpoi tt AMAYMi>-l 

aAi-iXY ' l »«'** w » 
14 t -|t:iXA( JIXMt • 
mi iijyi tMiitiy'it* 

Ml- l^rt.»MK*Jrypi J 

tt" |COIM'i»W"»^i 

KAioy^t hay r»* 

r:yi aih XT"' 1 *''* ' 


tlKI AOH ■llli't'Kf'l , t. 
miHir-JI II XI IAt»P ' 

id>i it i riiHiKii'V 

UMUiNdfNl t* If- 

XKoyi'Mtvi iuk* 


Miji Afi y^Nhi:i> 

e ixxv hit? AYTor+l- 


u>n AooMoyrXf h 

t'XNMXKr^WI" ■ 
r iorui>u>M AiAK- 
ilUihhtYftJhlTe'-'T- - 


IAC I 1 1 »yci hi AMOfK 





hi e tkai mm itrr-"" 


ACAirNY** 1 

t*N tn i loyhi n m-h 


cr ntox i K*ryo m 

i K'lllNM IN'MFT 


UDNON |tl»h*Oy 
A I yoh-ai I "tStAtT* 
oioai vivaI"*-"!" 

r iAUH)Y A 't: M ' ei * 

fl6TACWT'' oN " 

oK^ec iiM*f x * 


• HiriONxrroMKw 

i U>CT«Y "|t)A«c HAH 

ifi roNt'^'Aiti r* u 

OHtll'f Ion W*'" 

TAiuci t rot ict h 

fit Kfti»Mt» It t>l N 
MfliriMf/.MAt I" 
ItDflM-.M l"J<"»r> 




i A6|«i>yTU>«f Al 
hftV-cytJi aacoti 
*J»i Annexe rei Ay 

' A-MOY- IlkMNM-'-. 

r c i xyrtaiVi m «»n 


m t Ac i (j i xyran**' 
KtXeyoi AXt'trn 

VIM AtDCl At It lAY 
Hill lOIMAlMf |A 


" I M U»hl I UJ AN Mi ' 

lAili'MCVAYH' 1 

»i liniiirrjwu 

Tfl IliHKAlvptA*'- 

MrKAiAert T iAY 

AAUCYn hill > f K*l- 

U.A.CI •'lAY'ltH*****' 

* COKTrO tt:hJt<l> 
ft?fOO«- , /.<J>HN f 


fAfl K!fK ItHli-lIT 

MMNtUNI tt >tU> 
niNOTl^AI |t>l p 

t incithixt it iat 
t:i||firA>h* l«X< 

- 'A* Mil 
t CM hi* M l'i». 

I'm lor i " 




At 1 I At I CI AyiUJ 
O fO« Ahl AYTON «* 

Au>m t'N) Neajftf 
xtom ah uif occ*" 
cynoi AxoAoyer 

*' J.M M*e Nt>YNOT 


AAi/A^unr 00 ^ 10 



t:K« royKt nit hv 
Ayr u»o rt"o 1 1 trr •»■ 


M< Nl'l Mtrtl>C«p*J(...,. 
M ATX>y lOLff 1 1 H - 
fti>MIU,rri.>n""* - 
AAlMlMt < t HN1I 

mat ryfiAAYioy 

MOAAAAciionit -H 
Olt>ArtMAH?ANI fA 
Jt1'|KlhA«>( Moy 

I Al PaJ'OMi MXXi 
KAli> ' 

■ rYAIt-XI-ll 



■ •f""*ri|**«s" / * J ' 

Figure > Code:: Sinaiticus (^, fourth century), British Library, London; last 
folio of the Gospel according to John (see p. 65), John 21.1-25. Actual size 
14 7 /6X 13% inches. 

The type of text witnessed by Sinaiticus belongs in general to the 
Alexandrian group, but it also has a definite strain of the Western 
type. 23 Before the manuscript left the scriptorium, it was reviewed by 
several scribes who did the work of a dtogdorrfq (corrector). Readings 
which they are responsible for introducing are designated in a critical 

23- Especially in John 1.1-8.39, where it is a leading witness of the 
western text. See Gordon D. Fee, "Codex Sinaiticus in the Gospel of John: 
A Contribution to Methodology in Establishing Textual Relationships," New 
Testament Studies 15 (1968-9), pp. 23-44. 

Important Witnesses to the Text of the New Testament 67 

apparatus by the siglum R a . At a later date (probably sometime about 
the sixth or seventh century), a group of correctors working at Cae- 
sarea entered a large number of alterations into the text of both the Old 
and New Testaments. These readings, designated by the siglum s ca or 
R cb , represent a thoroughgoing effort to correct the text to a different 
standard, which, according to a colophon at the end of the books of 
Esdras and Esther, is stated to have been "a very ancient manuscript 
that had been corrected by the hand of the holy martyr Pamphilus." 24 

A. This handsome codex, dating from about the fifth century, 
contains the Old Testament, except for several mutilations, and most 
of the New Testament (the whole of Matthew's Gospel as far as 25.6 
is lost, as well as the leaves that originally contained John 6.50-8.52 
and 2 Cor. 4.13-12.6). It was presented in 1627 by Cyril Lucar, patri- 
arch of Constantinople, to King Charles I of England. Today, it rests 
along with Codex Sinaiticus in one of the prominent showcases in 
the Department of Manuscripts of the British Library. A photographic 
reproduction of the codex was published in 1879-83 by the British 
Museum, under the supervision of E. Maunde Thompson. Subse- 
quently, F. G. Kenyon edited a reduced facsimile of the New Testa- 
ment (1909) and of parts of the Old Testament. 

The quality of the text preserved in Codex Alexandrinus varies 
in different parts of the New Testament. In the Gospels, it is the old- 
est example of the Byzantine type of text, which is generally re- 
garded as an inferior form. In the rest of the New Testament (which 
may have been copied by the scribe from a different exemplar from 
that employed for the text of the Gospels), it ranks along with B and 
K as representative of the Alexandrian type of text. 25 

B. One of the most valuable of all the manuscripts of the Greek 
Bible is Codex Vaticanus. As its name indicates, it is in the great 
Vatican Library at Rome, which has been its home since some date 

24. Pamphilus, a native of Berytus (modern Beirut), devoted many 
years to hunting for and obtaining possession of books illustrative of the 
Scriptures from all parts of the world. His library, which was especially rich 
in biblical codices, was catalogued by Eusebius himself (Hist. Eccl. vi.32). 
Among its priceless treasures were the Hexapla and the Tetrapla of Origen 
in the original copies. 

25. H. Nordberg found that Codex Alexandrinus agrees with the 
dominant type of text represented in the biblical quotations made by 

St. Athanasius ("The Bible Text of St. Athanasius," Arctos, acta pbilologica 
Fennica, n.s. iii [1962], pp. 119-41). 

68 The Text of the New Testament 

prior to 1475, when it was mentioned in the first catalogue made of 
the treasures of the library. For some reason that has never been fully 
explained, during a large part of the nineteenth century, the author- 
ities of the library put continual obstacles in the way of scholars who 
wished to study it in detail. It was not until 1889-90 that a photo- 
graphic facsimile of the whole manuscript, edited by Giuseppe 
Cozza-Luzi, made its contents available to all. Another facsimile edi- 
tion of the New Testament was issued at Milan in 1904. 26 

The manuscript was written about the middle of the fourth cen- 
tury and contained both testaments as well as the books of the Apoc- 
rypha, with the exception of the books of Maccabees. Today, there 
are three lacunae in the codex: at the beginning almost 46 chapters 
of Genesis are missing, a section of some 30 Psalms is lost, and 
the concluding pages (from Heb. 9-14 onward, including 1 and 2 
Timothy, Titus, Philemon, and Revelation) are gone. 

The writing is in small and delicate majuscules, perfectly simple 
and unadorned. Unfortunately, the beauty of the original writing was 
spoiled by a later corrector, who traced over every letter afresh, omit- 
ting only those letters and words that he believed to be incorrect. 
The complete absence of ornamentation from Vaticanus has gener- 
ally been taken as an indication that it is slightly older than Codex 
Sinaiticus. On the other hand, some scholars believe that these two 
manuscripts were originally among the 50 copies of the Scriptures 
that the Emperor Constantine commissioned Eusebius to have writ- 
ten (see pp. 15-6). Indeed, T. C. Skeat of the British Museum has 
suggested that Codex Vaticanus was a "reject" among the 50 copies, 

26. In 1965, the New Testament portion of Codex Vaticanus was pho- 
tographically reproduced by order of Pope Paul VI and copies were pre- 
sented to the members and observers of Vatican Council II. The title page 
reads as follows: xa tega fiifiha I Codex Vaticanus graecus I 1209. I Pho- 
totypice expressus I iussu I Pauli PP VI I Pontificis Maximi I H KAINH 
AIA0HKH. I In Civitate Vaticana I 1965. The several forms of the edition 
contain also a second title page as well as an introduction of 21 pages 
(signed by Mgr. Paul Canart and Carlo M. Martini, S.J.) in English, French, 
German, Italian, and Spanish. 

On 25 Dec. 1999, the entire manuscript was issued in digitized format 
by Instituto poligrafico e zecca della stato, accompanied by a separate fas- 
cicle on paleographic and codicological matters (by Paul Canart), on the 
text of the Old Testament (by P.-M. Bogaert), and on the New Testament 
(by Stephen Pisano). 

Important Witnesses to the Text of the New Testament 69 

for it is deficient in the Eusebian canon tables, has many corrections 
by different scribes, and, as mentioned above, lacks the books of 
Maccabees, apparently through an oversight. Whether a "reject" or 
not, however, the text has been regarded by many scholars as an 
excellent representative of the Alexandrian type. 

In common with other manuscripts of the New Testament, the 
text of Vaticanus is divided into what may be called chapters. The 
system of division, however, appears to be more ancient than that 
current in other early parchment copies of the New Testament. For 
example, in the Epistles no notice is taken of 2 Peter; therefore, the 
system of divisions appears to date from a time when this epistle was 
not regarded as canonical. Furthermore, the Epistles of Paul have 
chapter numbers that do not begin afresh with each epistle (as is cus- 
tomary in other manuscripts) but continue in one series from Romans 
onward. In this manuscript, the Epistle to the Hebrews follows the 
Thessalonian Epistles; nevertheless, the sequence of the chapter 
numbers discloses that in an ancestor of Vaticanus Hebrews stood 
immediately after Galatians (cf. the sequence of epistles in p 46 ). 27 

C. Codex Ephraemi is the name given to a fifth-century manu- 
script of the Greek Bible that, as was mentioned in Chapter 1, was 
erased during the twelfth century and many of its sheets rewritten 
with the text of a Greek translation of 38 ascetical treatises or sermons 
by St. Ephraem, a Syrian Church father of the fourth century. By ap- 
plication of chemical reagents and painstaking labor, Tischendorf 
was able to decipher the almost totally obliterated underwriting of 
this palimpsest. 28 Only 64 leaves are left of the Old Testament, and of 

27. In Codex Vaticanus, the Epistle to the Galatians concludes with 
the 58th chapter, whereas the next epistle, that to the Ephesians, com- 
mences with the 70th chapter, and then the numbers continue regularly 
through Philippians, Colossians, and 1 and 2 Thessalonians, ending with 
the 93rd chapter. Following 2 Thessalonians stands Hebrews, which begins 
with the 59th chapter and proceeds with the 60th, 61 st, 62nd, 63rd, and 
64th chapters, as far as Heb. 9-14, where the manuscript breaks off, the re- 
maining pan being lost. It is clear from the sequence of chapter divisions 
that in an ancestor of Vaticanus I lebrews stood after Galatians and that the 
scribe of Vaticanus copied mechanically the chapter numerals even though 
they no longer were appropriate after Galatians. 

28. A list of corrections of Tischendorf s edition (Leipzig, 1843), based 
on a fresh examination of the manuscript, was published by Robert W. 
Lyon in New Testament Studies, v (1958-9), pp. 266-72. 

70 The Text of the N'kw Testament 

the New Testament there are 145 leaves (about five-eighths of the 
number that must have been originally required), containing portions 
of every book except 2 Thessalonians and 2 John. 

Though the document dates from the fifth century, its text is of 
less importance than one might assume from its age. It seems to be 
compounded from all the major text types, agreeing frequently with 
secondary Alexandrian witnesses but also with those of the later 
Koine or Byzantine type, which most scholars regard as the least 
valuable. Two correctors, referred to as C 2 or C h and C 3 or C c , have 
made corrections in the manuscript. The former probably lived in 
Palestine in the sixth century, and the latter seems to have done his 
work in Constantinople in the ninth century. 

D. Different in several respects from all the manuscripts men- 
tioned above is Codex Bezae (also known as Codex Cantabrigiensis), 
which was presented in 1581 to the library at Cambridge University 
by Theodore Beza, the celebrated French scholar who became the 
successor of Calvin as leader of the Genevan Church. Dating from 
the fifth century, this codex contains most of the text of the four 
Gospels and Acts, with a small fragment of 3 John. The text is pre- 
sented in Greek and Latin, the two languages facing each other on 
opposite pages, the Greek being on the left and the Latin on the 
right. Each page contains a single column of text, which is not writ- 
ten as a block but divided into sense-lines, that is, lines of varying 
length with the object of making the pauses in sense come at the 
end. The Gospels stand in the so-called Western order, with the two 
apostles first and the two companions of the apostles following 
(Matthew, John, Luke, and Mark). In each book, the first three lines 
are in red letters, and red ink and black ink alternate in the lines of 
the subscriptions at the end. In 1864, F. II, Scrivener published a 
careful edition of the manuscript, with full annotations; 29 and in 
1899, Cambridge University Press issued a handsome facsimile 
reproduction of the entire manuscript. The same publisher more 
recently issued what will, without doubt, remain for many genera- 
tions the definitive codicological study of the manuscript. David C. 
Parker of Birmingham, England, sets out his material in five parts: the 
paleography, the scribe and tradition, the correctors, the bilingual 
tradition, and the origins and history of the text. 30 

29. In 1978, the Pickwick Press of Pittsburgh published a reprint of 
Scrivener's volume. 

30. D, C. Parker, Codex Bezae: An Early Christian Manuscript and Its 
Text (Cambridge, 1992). 

Important Witnesses to the Text of the New Testament 71 

No known manuscript has so many and such remarkable varia- 
tions from what is usually taken to be the normal New Testament 
text. Codex Bezae's special characteristic is the free addition (and 
occasional omission) of words, sentences, and even incidents. Thus, 
in Luke 6, this manuscript has verse 5 after verse 10 and between 
verses 4 and 6 it contains the following account: "On the same day, 
seeing one working on the Sabbath day, he [Jesus] said to him, 'Man, 
if you know what you are doing, you are blessed; but if you do not 
know, you are accursed and a transgressor of the law' " (Fig. 10). 3I 
Although this sentence, which is found in no other manuscript, can- 
not be regarded as part of the original text of Luke, it may well em- 
body a first-century tradition, one of the "many other things that Jesus 
did" but that were not written in the Gospels (see John 21.25). In the 
Lucan account of the Last Supper (22.15-20), this manuscript (along 
with some Latin and Syriac witnesses) omits the latter part of verse 19 
and the whole of 20, thus removing all mention of the second cup 
and leaving the order of institution inverted (cup-bread). In Luke 
23.53, there is the additional information that Joseph of Arimathea, 
after laying the body of Jesus in his rock-hewn tomb, "put before the 
tomb a [great] stone that twenty men could scarcely roll." 

Codex Bezae is the principal authority, being supported by one 
other majuscule, <t>, the Old Latin and Curetonian Syriac versions, 
and a few copies of the Vulgate, in inserting after Matt. 20.28 the fol- 
lowing long passage: 

But seek to increase from that which is small, and to become less from 
that which is greater. When you enter into a house and are summoned 
to dine, do not sit down at the prominent places, lest perchance a man 
more honorable than you come in afterwards, and he who invited you 
come and say to you, "Go down lower"; and you shall be ashamed. But 
if you sit down in the inferior place, and one inferior to you come in, 
then he that invited you will say to you, "Go up higher"; and this will 
be advantageous for you. 32 

31. T. Nicklas discusses the interaction between the agraphon and its 
context in Codex Bezae ("Das Agraphon vom 'Sabbatarbeiter' und sein 
Kontext," Novum Testamentum, xliv [2002], pp. 160-75). 

32. English translations of Codex Bezae have been published by 
William Whiston ( The Primitive New Testament [London, 1745]) and by 
Johannes Greber ( The New Testament: A New Translation and Explanation 
Based on the Oldest Manuscripts [New York, 1937]). J. M. Wilson published 
The Acts of the Apostles, Translated from the Codex Bezae with an Intro- 
duction on its Lucan Origin and Importance (London, 1923). 


WurfrOHi flf«Wf n» « l» t tm « 

'hjI •Nri'irrrcviUir*' M»rr 

■2r .-..-r}.. I TilmMrti ttyi« «ln « wl >r>T 

d«i (u<H4ar(u(»l H-V>' 

- •.'*«- *H*Mar jlai wrMik* •»•»•*»«<•» 

■ « ,Uirtl**l«T1UlfUI.H|»Mt 
1U>fc<l|NIMl KTkAl Wl» » ^«J€-»>«-»»li U i< l »liy 
r*€-l KUOt t*>ur *•<*••*«»««•■*•» 

nu r m4mi •mwm an •#** «m w clac »f M • 

h» at u»«? X i»AMT»wiK«n[civ tr^Jr^'f T Mf 
t TTrAbA«lC-AT*«lr«lT 

I i < jii DHritiijrTiTt rumiM 

A*i<liMntuJ>r n(ti»Miai otij »»••*•>*►• r 
# uirvTCf itt *TrhA» t\ »»• i 
TIITAtlbAlnCUfAffr lHCN"fHl»fNI 
Atcntwnxx » t*A«i»»>!< '»»** 
rsciMi WN««l>'in*<'T* > 

wmrdw wttii«MV X»« *•«■ 
durtrAuirmHt Ad*«W w»r*»»«".'A»'«» 

».OiV VHl« I VAM1AIO ktM ..»,„»"„ 

ah i r» ><'«>' a«MIiia«-»»i »»••■-»* 


STm aSS^ "ir UryX UniVCtSiIy Ubn,ry ' Cambrid8e; Luke "^ (with an a * raph " n <**»»■ '- * 

Important Witnesses to the Text of the New Testament 73 

It is particularly in the Acts of the Apostles that Bezae differs 
markedly from other witnesses, 33 being nearly one-tenth longer than 
the text generally received. Thus, in Acts 12.10, this manuscript refers 
to the seven steps from the prison where Peter was kept down to the 
street. In 19-9, it adds the detail that at Ephesus Paul preached daily 
in the lecture hall of Tyrannus "from eleven o'clock to four," that is, 
at that time of day when the rhetorician Tyrannus would normally 
not hold his own sessions. In the decree that the Apostolic Council 
at Jerusalem issued about A.D. 50, this manuscript omits from the list 
of four prohibitions the clause "and from what is strangled" and adds 
at the close (Acts 15.20 and 29) the negative Golden Rule. 

These examples will be sufficient to indicate the characteristic 
freedom of the Western text, of which Codex Bezae is the principal 
representative. More study has been expended upon this manu- 
script, particularly where the Greek text differs from the parallel 
Latin text and where one or both differ from other witnesses, than 
has been devoted to any other New Testament witness. 34 There is 
still no unanimity of opinion regarding the many problems that the 
manuscript raises. 35 

D p . The symbol D p (sometimes called D2) refers to the sixth- 
century Codex Claromontanus, which contains only the Pauline 
Epistles (including Hebrews). Like Codex Bezae (which lacks the 
Pauline Epistles), D p is a bilingual Greek and Latin manuscript, hav- 
ing the Greek on the left-hand page and the text arranged in lines of 
irregular length corresponding to the pauses in the sense. The work 
of at least nine different correctors has been identified; the fourth of 

33. A history of research on Codex Bezae was compiled by Kenneth 
E. Panten, whose (unpublished) doctoral dissertation, entitled "A History 
of Research on Codex Bezae, with Special Reference to the Acts of the 
Apostles: Evaluation and Future Directions," was accepted by Murdock 
University, Perth, Australia, in 1995. A copy of the dissertation is also held 
by the Australian University in Canberra. 

34. See, e.g., A. R J. Klijn, A Survey of the Researches into the Western 
Text of the Gospels and Acts (Utrecht, 1949), continued in Novum Testa- 
mentum, iii (1959), pp. 1-27, 161-74. Happily, there is a complete Concor- 
dance to the Distinctive Greek Text of Codex Bezae, compiled by James D. 
Yoder (Grand Rapids, MI, 1961). 

35. See the essays in Christian-Bernhard Amphoux and David C. 
Parker, eds., Codex Bezae: Studies from the Lunel Colloquium, June 1994 
(Leiden, 1996). 

74 The Text of the New Testament 

these added accent and breathing marks in the ninth century. Like 
Codex Bezae, the type of text in this manuscript is distinctly Western; 
Western readings in the Epistles, however, are not so striking as 
those in the Gospels and Acts of Codex Bezae. An edition of the 
manuscript was published by Teschendorf in 1852. 

E. Codex Basiliensis, dating from the eighth century, contains 
the four Gospels on 318 leaves. It is now, as its name indicates, in 
the library of the University of Basle, Switzerland. It has a Byzantine 
type of text (see Fig. 6, p. 37). 36 

E a (also called E2). Formerly in the possession of Archbishop 
Laud, Codex Laudianus 35 of the Bodleian Library at Oxford dates 
from the late sixth century. It contains the Book of Acts in Latin and 
Greek, arranged in very short lines of only one to three words each, 
with the Latin in the left-hand column. The text exhibits a mixture of 
types, sometimes agreeing with Codex Bezae but more often with 
the Byzantine type. It is the earliest known manuscript that contains 
Acts 8.37 (the Ethiopian's confession of faith, see Fig. 11). An edition 
of the manuscript was published by Tischendorf in 1870. 

E p (also called E3). Codex Sangermanensis, now in St. Petersburg, 
contains the Pauline Epistles in Greek and Latin on opposite pages. It 
was copied in the ninth or tenth century from Codex Claromontanus 
and, therefore, is of no independent value. 

F. Codex Boreelianus, since 1830 in the University Library of 
Utrecht, once belonged to Johannes Boreel, Dutch ambassador at 
the court of James I of England. It contains the four Gospels (with 
large lacunae) and dates from the ninth century. Its text is typically 

F p (also called F2). Codex Augiensis, of the ninth century, con- 
tains the Pauline Epistles in double columns of Greek and Latin 
(Hebrews is given in Latin only). The manuscript once belonged to 
the monastery of Reichenau near Constance, which was known as 
Augia Maior; today, it is in the library of Trinity College, Cambridge. 
Its text, which was published by F. H. A. Scrivener in 1859, is of the 
Western type. 

G. Codex Wolfii A, also called Codex Harleianus, of the ninth 
century, contains the four Gospels with many lacunae. It was brought 

36. On Codex E, see Russell Champlin, Family E and Its Allies in 
Matthew (Studies and Documents, xxviii; Salt Lake City, UT, 1967), and 
Jacob Geerlings, Family E and Its Allies in Mark (Studies and Documents, 
xxxi; Salt Lake City, UT, 1968). 

Important Witnesses to the Text of the New Testament 75 

(3iiiuti;hti£ ooyuoy or. 

bAJ 1 VI '£*!«• KXI ri'iOOMMAIi 

(JuMrjunvtn (-nil n\t; 

t?i xy riu •;• 

P^»f,lffi o<|>i\iiiii«j. : . 

r\l'UTi) HXomu; 

inr.l.pi.f'1 ««cn i-icthiam 

piLni«n •iciivioii 

vNl'l TON,-, 

I? Till • K'AI l-| v t- vOyr^ 

&»Kt nip 'PU*f*M 

lVl'i> > -SO< H*VfUIM IWIKA'M KlMAH 


III* |UA<1» ^K'lTt.MIM' 

|'Jl»i.r/'|»U IfUOtfU* 1 l)Ml >tMIM»««« 

Figure 11 Codex Laudianus 35 (E a , sixth or seventh century), Bodleian 
Library, Oxford; Acts 8.36-8 (the earliest known witness to verse 37, see 
p. 74). Actual size WA X 8"A inches. 

from the East by Andrew E. Seidel in the seventeenth century and 
acquired by J. C. Wolf, who published extracts of it in 1723. Later, it 
became part of the library of Robert Harley and is now in the British 
Library. Its text is Byzantine. 

G p (also called G3). Codex Boernerianus, once owned by the 
Leipzig professor C. F. Borner, is now at Dresden. Dating from the 
ninth century, it contains the Pauline Epistles in Greek, with a liter- 
alistic Latin translation between the lines. After Philemon there 

76 The Text oe the New Testament 

stands the superscription for the Epistle to the Laodiceans, but the 
text of this apocryphal epistle is not present. In type of text it is 
closely akin to F p , and both of them probably go back one or two 
generations to a common archetype. 37 In many respects, it resembles 
the St. Gall MS. A and is thought to have been written in the 
monastery of St. Gall by some of the Irish monks who emigrated to 
those parts. At the foot of one of the leaves (fol. 23) are eight lines 
of Irish verse that refer to making a pilgrimage to Rome: 

To come to Rome, [to come to Rome,! 

Much of trouble, little of profit, 

The thing thou seekest here, 

If thou bring not with thee, thoufindest not? 6 

H. Codex Wolfii B was brought from the East with G and passed 
with it into the possession of J. C. Wolf. Its subsequent history is 
unknown until 1838, when it was acquired by the public library at 
Hamburg. Dating from the ninth century, it contains the four 
Gospels, with many lacunae. The text is of the Byzantine type. 

H a (also called H 2 ). Codex Mutinensis, a ninth-century copy of 
Acts (lacking about seven chapters), is in the Grand Ducal Library at 
Modena. The text is of the Byzantine type. 

H p (also called H 3 ). Codex Coislinianus is an important codex of 
the Pauline Epistles written in a very large hand with only a few 
words in each line. The text is Alexandrian. Dating from the sixth 
century, it came into the possession of the monastery of the Laura on 
Mount Athos, where, after it became dilapidated, its leaves were 
used to supply materials for the binding of several other volumes. 
Forty-one leaves are known to exist today, divided among libraries 
at Paris, St. Petersburg, Moscow, Kiev, Turin, and Mount Athos. A 
note appended to the Epistle to Titus states that it was corrected from 
the copy in the library of Caesarea, written by the hand of the holy 
Pamphilus himself. The text is arranged according to the colometri- 
cal edition of the Epistles prepared by Euthalius (or Evagrius), which 
is found in several other manuscripts (see p. 40). 

37. See W. H. P. Hatch, "On the Relationship of Codex Augiensis and 
Codex Boernerianus of the Pauline Epistles," Harvard Studies in Classical 
Philology, Ix (1951), pp. 187-99. 

38. For the entire text of the two stanzas, see F. H. A. Scrivener, A 
Plain Introduction to the Criticism of the New Testament, 4th ed., i (1894), 
p. 180, n. 2. 

Important Witnesses to the Text of the New Testament 77 

I. The Washington manuscript of the Pauline Epistles, in the 
Freer Museum at the Smithsonian Institution, originally contained 
about 210 leaves, of which only 84 survive in fragmentary condition. 
Dating from the fifth century, it contains portions of all the Pauline 
Epistles except Romans (the Epistle to the Hebrews follows 2 Thes- 
salonians). The text, which was edited by H. A. Sanders in 1921, is a 
good representative of the Alexandrian group, agreeing more closely 
with N and A than with B. 

K. Codex Cyprius, dating from the ninth century, is a complete 
copy of the four Gospels, with a typically Byzantine type of text. 

K ap (also called K 2 ). Codex Mosquensis is a ninth-century manu- 
script of Acts, the Catholic Epistles, and the Pauline Epistles (includ- 
ing Hebrews). The text, which is written in majuscule script, is sep- 
arated into paragraphs by comments, written in minuscule script. At 
the foot of the page are scholia, attributed to John Chrysostom. The 
text is a form of von Soden's /text (see p. 187). 

L. Codex Regius is an eighth-century codex of the Gospels, 
nearly complete, now in the Bibliotheque Nationale at Paris. It was 
edited by Tischendorf in 1846. Though badly written by a scribe who 
committed many ignorant blunders, its type of text is good, agreeing 
frequently with Codex Vaticanus (B). Its most notable feature is the 
presence of two endings to the Gospel according to Mark. The sec- 
ond of these is the traditional verses 9-20, but the first is a shorter 
ending, which is also found in a small number of other witnesses. 
This shorter ending reads as follows: "But they [the women] reported 
briefly to Peter and those with him all that they had been told. And 
after this, Jesus himself sent out by means of them, from east to west, 
the sacred and imperishable proclamation of eternal salvation." 

L ap (also called L2). Codex Angelicus, now in the Angelican Library 
at Rome, is a ninth-century copy of Acts, the Catholic Epistles, and the 
Pauline Epistles. Its text is mainly Byzantine. 

M. Codex Campianus, containing the four Gospels, is now in 
the Bibliotheque Nationale at Paris. It dates from the ninth century 
and contains a text that is mainly Byzantine but with admixture of 
Caesarean readings as well. 

N. One of the deluxe parchment manuscripts is Codex Pur- 
pureus Petropolitanus, written in the sixth century in silver letters on 
purple vellum, with gold ink for the contractions of the names of 
God and Jesus. Originally containing the four Gospels on approxi- 
mately 462 leaves, it was dismembered about the twelfth century, 
possibly by crusaders, and its leaves were carried far and wide. 

f NYV^T^** - •?*»" ^ • ' 

<W» •♦< f «i f) T*" « f* (pi 

T»f^j»r«^» W^ air. \ilf, 

f* neTpoviniOTK 

' ' '8KK*itT0icnVfetiiiH 
ajlaio li vnnof wicnoM 
TOvrOvVrivou IottA. 
•v MtV<M<vr vnf irum 




► VMTir.xioMlJuVTOClY 

- / . • VJ, '• 

)f* JfVf ,f - v x.UMI<lieif H 

I Tjrfy-iejiVirUUfvrvfY-wrt 

«K»vUJ-pUr- yVf jouV^y 
p* y6\»— w^i «p^Jf tfjrffl 


-jtUvVvt«v%xmc XmV. 
rtMNHeir m *>v.c eice\ 

UU4^.r T ^f l «^w te jyt.fl 

Figure 12 Codex Mosquensis (K ap , ninth century), State Historical Museum, 
Moscow; 1 Pet. 1.1-3, with commentary and scholia (see pp. 41 and 43). 
Actual size 13'/! X 9'/S inches. 


Important Witnesses to the Text of the New Testament 79 

Today, 182 leaves are in the Imperial Library at St. Petersburg; 33 
at Patmos; six in the Vatican Library; four in the British Library; two 
at Vienna; one in private possession at Lerma, Italy; one in the 
Byzantine Museum at Athens; and one in the Pierpont Morgan 
Library at New York. 39 The text belongs predominantly to the Byzan- 
tine type, but it preserves a number of readings of earlier types; B. 
H. Streeter regarded it (along with three other purple manuscripts, S, 
O, and 0) as a weak member of the Caesarean type. 40 

O. Codex Sinopensis is a deluxe edition, written in the sixth 
century with gold ink on purple vellum, of which 43 leaves of the 
Gospel according to Matthew survive (mainly chapters 13-24), as 
well as five miniatures. It was acquired at Sinope in Asia Minor by a 
French officer in 1899 and is now in the Bibliotheque Nationale at 
Paris. Its text, which was edited by Henri Omont in 1901, is a tertiary 
witness to the Caesarean type. 

P ;,pr (also called P 2 ). Codex Porphyrianus, now in St. Petersburg, 
a palimpsest dating from the ninth century, is one of the very few 
majuscule manuscripts that include the Book of Revelation (see n. 80, 
p. 51). In addition, it contains the Book of Acts and the Catholic and 
Pauline Epistles, though with lacunae. The upper writing, which is 
dated in the year 1301, consists of the commentary of Euthalius on the 
Acts and the Pauline Epistles, together with the biblical text. 
Tischendorf edited the manuscript in 1865-9- According to von Soden, 
in Acts the text is Koine with sporadic /readings and in the other 
books it is Alexandrian. According to Schmid, however, in Revelation 
the text is a secondary development of a basic Andreas type. 

R. Codex Nitriensis, now in the British Library, is a palimpsest 
containing parts of Luke in a fine large hand of the sixth century, 
over which the Syriac treatise of Severus of Antioch against Johannes 
Grammaticus was written in the eighth or ninth century. The same 
manuscript also contains a palimpsest of 4,000 lines of Homer's 
Iliad. Along with more than 500 other manuscripts, it was brought to 
England in 1847 from the monastery of St. Mary the Mother of God, 
in the Nitrian Desert, 70 miles northwest of Cairo. According to von 

39. The last two leaves (containing Matt. 14.22-31 and 15.38-16.7) 
were edited by Stanley Rypins in the Journal of Biblical Literature, lxxv 
(1956), pp. 27-39. 

40. B. H. Streeter, "Codices 157, 1071 and the Caesarean Text," Quan- 
tulacumque, Studies Presented to Kirsopp Lake (London, 1937), pp. 149-50. 

80 The Text oe the New Testament 

Soden, the text belongs to his / (i.e., Western) type. Tischendorf 
edited the text in 1857. 

S. This is one of the earliest dated Greek manuscripts of the 
Gospels; a colophon states that it was written by a monk named 
Michael in the year of the world 6457 (= A.D. 949). It is now in the 
Vatican Library (no. 354). The text type is Byzantine. 

T. Codex Borgianus, at the Collegium de Propaganda Fide in 
Rome, is a valuable Greco-Sahidic manuscript of the fifth century.' 11 
Unfortunately, it is fragmentary, preserving only 179 verses of Luke 
22-23 and John 6-8. The text is very close to that represented by 
Codex Vaticanus (B). 

V. Codex Mosquensis, formerly of the monastery of Vatopedi on 
Mount Athos and now in Moscow, is a copy of the four Gospels, 
nearly complete, dating from the ninth century. The manuscript is 
written in majuscules down to John 8.39, where it breaks off, and 
from that point the text is continued in a minuscule hand of the thir- 
teenth century. The text type is Byzantine (Fig. 12, p. 78). 

W. Among the more important majuscule manuscripts discov- 
ered during the twentieth century is a codex of the four Gospels ac- 
quired by Charles L. Freer of Detroit in 1906 and now in the Freer 
Museum of the Smithsonian Institution in Washington, D.C. It dates 
from the late fourth or early fifth century and, like Codex Bezae, con- 
tains the Gospels in the so-called Western order (Matthew, John, 
Luke, and Mark). Each of the two leaves that serve as covers has two 
painted panels depicting two of the Evangelists; these miniatures are 
thought to date from about the seventh or eighth century. The type 
of text is curiously variegated, as though copied from several manu- 
scripts of different families of text. In Matthew and Luke 8.13-24.53, 
the text is of the common Byzantine variety; but in Mark 1.1-5.30, it 
is Western, resembling the Old Latin; Mark 5.31-16.20 is Caesarean, 
akin to p 45 ; 42 and Luke 1.1-8.12 and John 5.12-21.25 are Alexandrian. 
The text of John 1.1-5.11, which fills a quire that was added about 
the seventh century, presumably to replace one that was damaged, 
is mixed, with some Alexandrian and a few Western readings. In the 
opinion of its editor, Henry A. Sanders, this stratification of different 

41. The text of the fragment of John was edited by A. A. Giorgi, Frag- 
mentum Evangelii S.Johannis Graecum Copto-Sahidicum . . .(Rome, 1789). 

42. For a full analysis of the text of W in Mark, see Larry Hurtado, 
Text-Critical Methodology and the Pre-Caesarean Text: Codex Win the 
Gospel of Mark (Grand Rapids, MI, 1981). 

Important Witnesses to the Text of the New Testament 81 

kinds of text is explained by the theory that the codex goes back to 
an ancestor made up of fragments from different manuscripts of the 
Gospels pieced together after the attempt of the Emperor Diocletian 
to crush Christianity by destroying its sacred books. 

One of the most noteworthy of the variant readings in Codex W 
is a remarkable insertion near the close of the Gospel according to 
Mark, part of which was known to Jerome, who declares that it was 
present "in certain copies and especially in Greek codices." Following 
the reference to the appearance of the risen Christ, who upbraided the 
1 1 "for their lack of belief and stubbornness of heart, because they 
had not believed those who saw him after he had risen" (Mark 16.14), 
the text proceeds immediately with the following addition: 

And they excused themselves, saying, "This age of lawlessness and un- 
belief is under Satan, who does not allow the truth and power of God 
to prevail over the unclean things of the spirits. 43 Therefore reveal your 
righteousness now" — thus they spoke to Christ. And Christ replied to 
them, "The term of years of Satan's power has been fulfilled, but other 
terrible things draw near. And for those who have sinned I was deliv- 
ered over to death, that they may return to the truth and sin no more, 
that they may inherit the spiritual and imperishable glory of righteous- 
ness that is in heaven. " Ai 

X. Codex Monacensis, now in the University Library at Munich, 
contains portions of the four Gospels in the order Matthew, John, 
Luke, and Mark. It dates from the tenth century. Except in Mark, the 
text is interspersed with a patristic commentary, which is written in 
a contemporary minuscule hand. Though its text is mainly of the 
Byzantine type, it also contains occasional readings of an earlier 
type, akin to Alexandrian. 

Z. Codex Dublinensis is an interesting palimpsest in the library 
of Trinity College, Dublin. It consists of 32 leaves and preserves 295 
verses of Matthew in large and broad majuscules of the sixth century. 
The text agrees chiefly with that of Codex Sinaiticus. The manuscript 
was edited by T. K. Abbott in 1880. 

43. Or "does not allow what lies under the unclean spirits to under- 
stand the truth and power of God." 

44. The text of this passage, with notes and commentary, was edited 
by Caspar Ren6 Gregory in Das Freer-f.ogion (Leipzig, 1908). See also Jorg 
Frey, "7m Text und Sinn des Freer-Logion," Zeitschrift fur neuWstamentlicbe 
Wissenschaft, xciii (2002), pp. 13-34. 

82 The Text of the New Testament 

v " Jc«U Jrte*rn»-T«c«nfc«««i<ifi-T«5cfpt«»r Jciaicajwc hi^uha- 

£*'■ l« «".«? i^tovv^ _ tfAt—r ,r"-''«-*«r i*fnf^.A- 

t.-o . Alum " *•>.»- J>r*fc»/- J «<*.'* cL"na«nr' In 

> apse jc.« ForHoc-y.itTrtsutdHct^r-u- rcr*'-.T-*-<* , «l<.* 

v JL*!. ^"MKI. . ' ,V -M-46* «" »t*r* pt..«.»r .£■ wit 

i ^fcl ^'' W -"'"nwA-**^ rtc *A0vc Cms. (c.Vi c'f ■«»■ 

*y- Vttr •wmr t i4^ :'- ,fr" ^ .»'«*<»-«•■ rt-w 

T<»T** <H«»oV*V*»WWv*. j^liHC- b.v«rr«*t«H.U yflArTO^ Tyr 


Figure 13 Codex Sangallensis (A, ninth century), Monastery of St. Gall, 
Switzerland; Greek text with Latin interlinear; Luke 2.51-3-7 (see p. 83). 
Actual size 8^ X7% inches. 

A. Codex Sangallensis is a ninth-century Greco-Latin manu- 
script, the Latin version being written between the lines of the Greek 
(Fig. 13). It contains the four Gospels complete, with the exception 
of John 19-17-35. In Mark, its text belongs to the Alexandrian type, 
similar to that of L; in the other Gospels, however, it belongs to 

Important Witnesses to the Text of the New Testament 83 

the ordinary Koine or Byzantine type. The manuscript was edited 
by H. C. M. Rettig in 1836. A curious mistake occurs at Luke 21.32 
where the scribe wrote d(t>daXfiOvq ("eyes") instead of ddsAipovg 

0. Codex Koridethi is a manuscript of the Gospels that was dis- 
covered in the church of SS. Kerykos and Julitta at Koridethi, located 
in the Caucasian Mountains near the Caspian Sea; it is now at Tbilisi, 
a city of the Republic of Georgia. is written in a rough, inelegant 
hand, by a scribe who clearly was not familiar with Greek. Its edi- 
tors, Gustav Beermann and C. R. Gregory, date the manuscript in the 
ninth century. In Matthew, Luke, and John, the text is similar to most 
Byzantine manuscripts, but in Mark it is quite different; here, it is 
akin to the type of text that Origen and Eusebius used in the third 
and fourth centuries at Caesarea. 

A. Codex Tischendorfianus III, now in the Bodleian Library at 
Oxford, contains the text of Luke and John in a ninth-century hand 
characterized by sloping Slavonic majuscules. Its text is mainly 
Byzantine. At the close of the Gospels stands the so-called Jerusalem 
colophon (see the description of MS. 157, pp. 88-9). 

E. One of the most interesting palimpsest manuscripts is Codex 
Zacynthius, a fragmentary codex preserving the greater part of Luke 
1.1-11.33. It was brought from the isle of Zante in 1821 and is today in 
the library of the British and Foreign Bible Society in Cambridge. It is 
the earliest known New Testament manuscript with a marginal com- 
mentary, and it is the only one that has both text and commentary in 
majuscule script. This commentary, which surrounds the single col- 
umn of text on three sides, is a catena of quotations from the exegeti- 
cal writings of nine Church fathers. 45 The type of text is Alexandrian, 
akin to that of Codex Vaticanus (B); and it has the same system of 
chapter divisions, which is peculiar to these two majuscule manu- 
scripts and to Codex 579. Written in the sixth century, it was erased in 
the twelfth or thirteenth century, and the sheets were reused to receive 
the text of a Gospel lectionary. It was edited by S. P. Tregelles in 1861. 

77. Codex Petropolitanus is a copy of the four Gospels, almost 
complete (it lacks 77 verses of Matthew and John). Dating from the 
ninth century, it contains a Byzantine type of text, being the head 

45. See J. H. Greenlee, "The Catena of Codex Zacynthius," Biblica, xl 
(1959), pp. 992-1001. See also D. C. Parker and J. Neville Birdsall, "The 
Date of Codex Zacynthius ," Journal of Theological Studies, n.s., lv (2004), 
pp. 115-31 ("around the year 700"). 

84 The Text of the New Testament 

of a subfamily that is akin to, but not descended from, Codex 
Alexandrinus. 46 

S. Codex Rossanensis, containing Matthew and iMark, is written 
on thin vellum stained purple, in silver letters, the first three lines of 
each Gospel being in gold. Dating from the sixth century, it is the 
earliest known copy of Scripture that is adorned with contemporary 
miniatures in watercolors, 17 in number. 47 These include the raising 
of Lazarus, the driving of the traders out of the temple, the ten vir- 
gins, the entry into Jerusalem, the foot-washing, the Last Supper, and 
Jesus before Pilate (Fig. 14). Its text (which was edited by O. von 
Gebhardt in 1883) is closely akin to that of N, agreeing frequently 
with the Byzantine type of text but with certain Caesarean readings 
as well. The manuscript belongs to the Archbishop of Rossano, at the 
southern end of Italy. 

0. Codex Beratinus of the sixth century is (like manuscripts N, 
O, and S) a deluxe purple vellum manuscript written with silver ink. 
It contains only Matthew and Mark, with several considerable lacu- 
nae, and is in the possession of the Church of St. George at Berat in 
Albania. Its text (which was edited by P. Batiffol in 1887) is generally 
of the Koine type, but it contains the long Western addition after 
Matt. 20.28, already quoted as occurring in D. According to Streeter, 
the manuscript is a tertiary witness to the Caesarean text. 

W. Codex Athous Laurae, as its name implies, is a manuscript in 
the monastery of the Laura on Mount Athos. Dating from the ninth or 
tenth century, it contains the Gospels (from Mark 9 onward), Acts, the 
Catholic Epistles (in the unusual order of Peter, James, John, and 
Jude), the Pauline Epistles, and Hebrews (except one leaf of the last). 
It agrees with L in giving the shorter ending of Mark before the longer 

46. On the textual affinities of Codex Petropolitanus, see Silva Lake, 
Family 17 and the Codex Alexandrinus; The Text According to Mark {Stud- 
ies and Documents, v; London, 1937); Jacob Geerlings, Family IJ in Luke 
{Studies and Documents, xxii; Salt Lake City, UT, 1962); idem, Family 77 in 
John {Studies and Documents, xxiii; Salt Lake City, UT, 1963); and Russell 
Champlin, Family TJ in Matthew {Studies and Documents, xxiv; Salt Lake 
City, UT, 1964). 

47. A study of the manuscript in its artistic aspects, with photographic 
reproductions of all the miniatures, was published by A. Haseloff, Codex 
Purpureus Rossanensis (Leipzig, 1898). Another reproduction, with the plates 
in color, was edited by A. Munoz, II codice purpurea di Rossano (Rome, 
1907), For a commentary (in Italian and English) on the miniatures, see 

G. Cavallo, J. Gribomont, and W. C. Loerke, Codex Purpureus Rossanensis, 
Commentarium (Rome, 1987), pp. 45—171. 

Important Witnesses to the Text of the Xew Testament 

» r 

Figure 14 Cod- r.ossana • Z sixth cenr ry), in the archives of the 
Archiepiscopal Curia, Rossano, Calabria; folio 8 v (Munoz), Christ and Barab- 
bas before Pilate (with the court stenographer standing by me table; see 
p. 28, n. 40). Actual size \l% X 9V» inches. 

one. According to Kirsopp Lake, 48 its text in Mark is an early one, with 
readings both Alexandrian and Western but chiefly akin to the group 
N, C, L, and A. The other Gospels are predominantly Byzantine, with 
a somewhat larger proportion of Alexandrian readings than in A. 

48. Lake published the text of Mark and a collation of I.uke, John, and 
Colossiaps in Studia Biblica et Ecclcsiastica, v (Oxford, 1W), pp. ^M-13 1 

86 The Tkxt of the New Testament 

Q. Codex Athous Dionysiou, a complete copy of the four Gospels 
(except Luke 1.15-28) in the monastery of Dionysius on Mount Athos, 
dates from the ninth century. Von Soden classed it as one of the three 
oldest manuscripts that, in his opinion, present the earliest variety of 
the Koine or Byzantine text. A collation made by Mary W. Winslow 
was published in 1932 by Kirsopp Lake and Silva New. 49 

046. Codex Vaticanus 2066, dating from the tenth century, con- 
tains the Rook of Revelation between treatises of Basil and Gregory' 
of Nyssa. Previously, it was designated B r or B2, which gave rise to 
confusion with the famous Codex Vaticanus (B). In text type, it is re- 
lated to minuscules 6l and 69, with a form of text that differs from 
the early majuscules as well as the later ecclesiastical text. 

0171. This numeral is given to two parchment fragments from 
Egypt dating from about 300 and containing Luke 21.43-7, 50-3 and 
22.44-56, 61-3. According to Lagrange, it is an important witness in 
Egypt to the Western text. 50 

0220. This third-century parchment leaf of Romans (4.5-5.3 and 
5.8-13) was purchased at Cairo in 1950 by Dr. Leland C. Wyman, 
professor of biology at Boston University. The importance of 0220 
lies in its agreement with Codex Vaticanus everywhere except in 5.1, 
where it apparently reads the indicative l^o/iev. 51 

3. Important Greek Minuscule Manuscripts 
of the New Testament 

The more important minuscule manuscripts of the New Testa- 
ment include the following. In several cases, scholars have discovered 
that certain manuscripts exhibit such striking similarities of text type 
as to suggest a close "family" relationship. 

Fam. 1 . Early in the twentieth century, Kirsopp Lake' 52 identified 
a family of witnesses that includes manuscripts 1, 118, 131, and 209, 
all of which date from the twelfth to the fourteenth centuries. Textual 
analysis of the Gospel according to Mark indicates that the type of 

49. Kirsopp Lake and Silva New, Six Collations of New Testament Man- 
uscripts (Harvard Theological Studies, xvii; Cambridge, MA, 1932), pp. 3-25. 

50. The text is edited in Pubblicazioni della Societd Italiana, Papiri 
Greci e Latini, i (Florence, 1912), pp. 2-4, and ii (1913), pp. 22-5. 

51. The leaf was edited by W. H. P. Hatch in the Harvard Theological 
Review, xlv (1952), pp. 81-5. 

52. In Texts and Studies, vii (2) (Cambridge, 1902). 

Important Witnesses to the Text of the New Testament 87 

text preserved in these minuscules often agrees with that of Codex 
and appears to go back to the type current in Caesarea in the third 
and fourth centuries. Recently, it has been argued that 1582 should 
be seen as the leading member of the group in Matthew. 

Fam. 13. In 1868, a professor of Latin at Dublin University, 
William Hugh Ferrar, discovered that four medieval manuscripts, 
namely 13, 69, 124, and 346, were closely related textually. His col- 
lations were published posthumously in 1877 by his friend and col- 
league, T. K. Abbott. It is known today that this group (the Ferrar 
group) comprises about a dozen members (including manuscripts 
230, 543, 788, 826, 983, 1689, and 1709). They were copied between 
the eleventh and fifteenth centuries and are descendants of an 
archetype that came either from Calabria in southern Italy or from 
Sicily. 53 One of the noteworthy features of these manuscripts is that 
they have the section about the adulterous woman (John 7.53-8.11), 
not in the fourth Gospel, but after Luke 21.38. Like fam. 1, this fam- 
ily also has affinities with the Caesarean type of text. 54 

MS. 28. This eleventh-century copy of the four Gospels (with la- 
cunae) is carelessly written but contains many noteworthy readings, 
especially in Mark, where its text is akin to the Caesarean type. It is 
in the Bibliotheque Nationale at Paris; a collation was published by 
the Lakes. ^ 

MS. 33. Since the time of J. G. Fichhorn in the early nineteenth 
century, MS. 33 has often been called "the queen of the cursives." 
Now in the Bibliotheque Nationale at Paris, it is an important minus- 
cule codex, containing the entire New Testament except the Book of 

53. See Robert Devreesse, les Manuscrits grecs de lltalie meridionale 
(histoire, classemenl, paleographie) (Studi e testi, clxxxiii; Citta del Vaticano, 

54. Kirsopp and Silva Lake, Family 13 (.The Ferrar Group) (Studies 
and Documents, xi; Philadelphia, 1941). See also Jacob Geerlings, The 
Lectionary Text of Family 13 According to Cod Vat Gr 121 7 ( Gregory 547) 
(Studies and Documents, xv; Salt Lake City, UT, 1959); idem, Family 13- 
The Ferrar Group. The Text According to Matthew (Studies and Documents, 
xix; Salt Lake City, UT, 1964); idem, Family 13 (The Ferrar Group). "The 
Text According to Luke (Studies and Documents, xx; Salt Lake City, UT, 
1961); idem, Family 13 ('The Ferrar Group). The Text According to John 
(Studies and Documents, xxi; Salt Lake City, UT, 1962). 

55. The collation was published in their monograph Family 13, 
pp. 117-54. 

88 The Text of the New Testament 

Revelation and dating from the ninth century. It is an excellent rep- 
resentative of the Alexandrian type of text, but it shows also the in- 
fluence of the Koine or Byzantine type, particularly in Acts and the 
Pauline Epistles. 

MS. 61. This manuscript of the entire New Testament, dating 
from the early sixteenth century, now at Trinity College, Dublin, has 
more importance historically than intrinsically. It is the first Greek 
manuscript discovered that contains the passage relating to the 
Three Heavenly Witnesses (1 John 5.7-8). It was on the basis of this 
single, late witness that Erasmus was induced to insert this certainly 
spurious passage into the text of 1 John. The manuscript, which is re- 
markably fresh and clean throughout (except for the two pages con- 
taining 1 John 5, which are soiled from repeated examination), gives 
every appearance of having been produced expressly for the pur- 
pose of confuting Erasmus (see p. 146). 

MS. 69. Containing the entire New Testament, this manuscript 
was copied in the fifteenth century by a Greek named Emmanuel, 
from Constantinople, who worked for Archbishop Neville of York 
about 1468. "^ Written partly on vellum and partly on paper, it is an 
important member of fam. 13 (edited by T. K. Abbott with other 
members of that family). The manuscript is now in the Guildhall 
museum of Leicester, England. 

MS. 81. Written in 1044, this manuscript, now in the British Li- 
brary, is one of the most important of all minuscule manuscripts. It 
contains the text of Acts in a form that agrees frequently with the 
Alexandrian type. It was collated by Scrivener. v 

MS. 157. This is a handsome codex of the Gospels, dating from 
about 1122, now in the Vatican Library, written for the Emperor 
John II Comnenus (reigned 1118-43). Its text resembles that of MS. 
33 and was thought by Streeter to belong to the Caesarean type." 18 

56. M. R. James, Journal of Theological Studies, v (1904), pp. 445-7; xi 
(1910), pp. 291-2; and xii (1911), pp. 465-6. See also idem, The Wander- 
ings and Homes of Manuscripts (London, 1919), pp. 17 f. Concerning the 
scribe Emmanuel, see H. I- Gray, "Greek Visitors to England in 1455- 
1456," in Anniversary Essays in Mediaeval History, by students of Charles 
Homer Haskins (New York, 1919), pp. 81-116, especially 105 ff. 

57. EH. Scrivener, An Exact Transcript of the Codex Augiensis . . . 
to which Is Added a Eull Collation of Fifty Manuscripts (Cambridge, 1859). 

58. Streeter, op. cit., pp. 149-50. 

Important Witnesses to the Text of the New Testament 89 

A colophon, which is also found in a dozen other manuscripts (A , 20, 
164, 215, 262, 300, 376, 428, 565, 686, 718, and 1071), states that it was 
copied and corrected "from the ancient manuscripts at Jerusalem." 
This colophon is repeated after each of the four Gospels. A collation 
of the manuscript was published by H. C. Hoskier. 59 

MS. 383. This is a thirteenth-century codex of Acts and the Epistles 
(Catholic and Pauline) in the Bodleian Library at Oxford. It was col- 
lated by August Pott for his volume Der abendldndische Text der 
Apostelgeschichte und die Wir-Quelle (Leipzig, 1900), pp. 78-88, and 
used by A. C. Clark in his reconstruction of the Western text of Acts. 

MS. 565. One of the most beautiful of all known manuscripts, 
565 is now in the public library at St. Petersburg. It is a deluxe copy 
of the Gospels, written in gold letters on purple vellum during the 
ninth century. 60 In Mark, it is an ally of in support of the Caesarean 
text. At the close of Mark, it contains the so-called Jerusalem 
colophon (see the description of Codex 157). 

MS. 579. This is a thirteenth-century copy of the Gospels in the 
Bibliotheque Nationale at Paris. 61 In Matthew, its text belongs to the 
late Byzantine group, but in the other Gospels it preserves an ex- 
tremely good Alexandrian text that often agrees with B, K, and L. Like 
MS. L, it contains the double ending of Mark. 

MS. 614. A thirteenth-century codex of Acts and the Epistles 
(Pauline and Catholic) from Corfu, MS. 6l4 is now in the Ambrosian 
Library at Milan. It contains a large number of pre-Byzantine read- 
ings, many of them of the Western type of text. 62 

MS. 700. This eleventh-century codex of the Gospels, now in the 
British Library, diverges 2,724 times from the Textus Receptus and has 
besides 270 readings peculiar to itself. 63 Along with one other Greek 

59. Journal of Theological Studies, xiv (1913), pp. 78 ff., 242 ff., 359 ff. 

60. The text of Mark and a collation of Matthew, Luke, and John were 
published by Johannes Belsheim in Christiania Videnskabs-Selskabs 
Forhandlinger, 1885, no. 9. Corrections of Belsheim's edition were included 
by H. S. Cronin in Texts and Studies, v (4) (Cambridge, 1899), pp. 106-8. 

61. Alfred Schmidtke published an edition of Mark, Luke, and John in 
his Die Evangelien eines alten Uncial codex (B^-Text) nach einer Abschrift 
des dreizehnten Jahrhunderts (Leipzig, 1903). 

62. An edition prepared by A. V. Valentine-Richards was published 
posthumously, with an introduction by J. M. Creed (Cambridge, 1934). 

63. A collation was published by H. C. Hoskier in A Full Account and 
Collation of the Greek Cursive Codex Evangelium 604 (London, 1890). 

90 The Text of the New Testament 

manuscript (MS. 162) it has the remarkable reading in the Lucan form 
of the Lord's Prayer, "May your Holy Spirit come upon us and cleanse 
us," instead of "May your kingdom come" (11.2). This was also the text 
of the Lord's Prayer known to Marcion and Gregory of Nyssa/' 4 

MS. 892. This is a ninth-century codex of the four Gospels, ac- 
quired by the British Museum in 1887. fc It contains many remarkable 
readings of an early type, belonging chiefly to the Alexandrian text. 
Von Soden observed that the scribe of 892 preserved the divisions in 
pages and lines of its majuscule parent. 

MS. 1071. This twelfth-century copy of the four Gospels, now 
in the Laura on Mount Athos, contains the so-called Jerusalem 
colophon referred to in the description of codex 157. 66 Streeter clas- 
sified its text as a tertiary witness to the Caesarean type. 

MS. 1241. This manuscript, containing the whole New Testament 
except the Book of Revelation, dates from the twelfth century. 67 In 
the Gospels, its text has some agreements with C, L, A, V, and 33- 
According to Kirsopp Lake, in Matthew and Mark its text shows a 
larger infusion of Byzantine readings than in Luke and John. 

Fam. 1424. Codex 1424 is a ninth- or tenth-century copy of the 
entire New Testament, written by a monk named Sabas in the se- 
quence of Gospels, Acts, Catholic Epistles, Revelation, and Pauline 
Epistles. All the books except Revelation are supplied with a com- 
mentary, which is written in the margins. Formerly in the monastery 
at Drama (Turkish Kosinitza) in Greece, it was probably taken 
thence to western Europe after the Balkan wars of 1912-13. It was 
bought by Dr. L. Franklin Gruber, president of the Chicago Lutheran 

64. For what can be said in support of the opinion that this variant 
reading stood in the original text of Luke, see Robert Leaney in Novum Tes- 
tamenlum, i (1956), pp. 103-11; for a statement of the view that this form 
of the Lord's Prayer represents a modification of the usual form for use at 
special services (such as ordination), see the discussion of B. M. Metzger in 
Twentieth Century Encyclopedia of Religious Knowledge, ii (Grand Rapids, 
MI, 1955), pp. 673 f. 

65. A collation was published by J. R. Harris in the Journal of Biblical 
Literature, ix (1890), pp. 31-59. 

66. A collation of Mark and of several chapters from the other 
Gospels was published by Kirsopp Lake in Studia Biblica et Ecclesiastica, v 
(Oxford, 1903), pp. 140-8. 

67. A collation of the Gospels was published by Kirsopp Lake, Six 

Important Witnesses to the Text of the New Testament 91 

Theological Seminary at Maywood, Illinois, and bequeathed at his 
death to the seminary library. According to von Soden, its text in the 
Gospels belongs to his /* group, which Streeter renamed fam. 1424 
and classified as a tertiary witness to the Caesarean text. In addition 
to MS. 1424, which is the oldest minuscule of the family, the other 
members of the family are M, 7, 27, 71, 115 (Matthew, Mark), 160 
(Matthew, Mark), 179 (Matthew, Mark), 185 (Luke, John), 267, 349, 
517, 659, 692 (Matthew, Mark), 827 (Matthew, Mark), 945, 954, 990 
(Matthew, Mark), 1010, 1082 (Matthew, Mark), 1188 (Luke, John), 
1194, 1207, 1223, 1293, 1391, 1402 (Matthew, Mark), 1606, 1675, and 
2191 (Matthew, Mark). 

MS. 1582. Written in 948 by the scribe Ephraim, manuscript 1582 
contains an ancient and valuable text of the Gospels, which has re- 
cently been recognized as potentially significant for understanding 
and reconstructing the readings of the Caesarean text. A recent study 
by Amy Anderson has shown that in the Gospel according to Matthew, 
1582 contains text and marginal notes that are closely related to the 
third-century text used by Origen and that the manuscript is best seen 
as the leading member of the manuscripts that comprise fam. I. 68 

MS. 1739. Containing the Acts and the Epistles, this tenth-century 
manuscript was discovered at Mount Athos in 1879 by E. von der 
Goltz and is usually known by his name. 69 It is of extreme impor- 
tance because it contains a number of marginal notes taken from the 
writings of Irenaeus, Clement, Origen, Eusebius, and Basil. Since 
nothing is more recent than Basil, who lived from A.D. 329 to 379, it 
appears that the ancestor of this manuscript was written by a scribe 
toward the close of the fourth century. A colophon indicates that for 
the Pauline Epistles the scribe followed a manuscript that contained 
an Origenian text. It is, however, not of the Caesarean type but pre- 
sents a relatively pure form of the Alexandrian type. 

MS. 2053. This is a thirteenth-century manuscript at Messina, 
containing the text of the Book of Revelation with Oecumenius' 
commentary on it. Along with Codices A, C, and 2344, it is (accord- 
ing to Schmid) 70 one of the best sources for the text of the Apoca- 
lypse, superior even to p 47 and K. 

68. Amy Anderson, The Textual Tradition of the Gospels; Family One in 
Matthew (Leiden, 2004), Chapter 6, "Codex 1582 and Codex 1," pp. 84-102. 

69. A collation was made by Morton S. Enslin, in Lake Six Collations. 

70. Josef Schmid, Studien zur Geschichte des griechischen 
Apokalypse-Textes: 2. Teil, Die alien Stdmme (Munich, 1955), p. 24. 

92 Ti ik Text of the New Testament 

MS. 2344. An eleventh-century codex now in the Bibliotheque 
Nationale at Paris, MS. 2344 contains Acts, Catholic Epistles, Pauline 
Epistles, and Revelation (where it agrees frequently with MS. 2053), 
besides parts of the Old Testament. 

4. Other Noteworthy Manuscripts 

Manuscripts that are noteworthy because of their external format or 
in some other respect include the following. A majuscule copy of the 
four Gospels, 71 047, dating from the eighth century and now in the 
Princeton University Library, has the writing on each page arranged 
in the form of a cross; that is, the lines comprising the top third of 
the column and those of the bottom third are about one-half the 
length of the lines comprising the middle section of the column. 

Codex 16, a fourteenth-century copy of the four Gospels in 
Greek and Latin, formerly in the possession of Catherine de Medici 
and now in the Bibliotheque Nationale at Paris, is written in four col- 
ors of ink according to the contents. The general run of the narrative 
is in vermilion; the words of Jesus, the genealogy of Jesus, and the 
words of angels are in crimson; the words quoted from the Old 
Testament as well as those of the disciples, Zachariah, Elizabeth, 
Mary, Simeon, and John the Baptist are in blue; and the words of the 
Pharisees, the centurion, Judas Iscariot, and the devil are in black. 
The words of the shepherds are also in black, but this may well have 
been an oversight. 

One of the smallest Greek manuscripts containing the four 
Gospels is MS. 461, now in the public library at St. Petersburg. There 
are 344 leaves, each of which measures 6% by 3% inches; the single 
column of writing occupies an area of about AVi by 2% inches. 
This manuscript is noteworthy also because it is the earliest dated 
Greek minuscule manuscript known to exist, having been copied in 
A.D. 835. 72 

Even more tiny must have been the vellum codex of the Book 
of Revelation, of which only one leaf is extant (MS. 0169, now in the 

71. A plate is included in B. M. Metzger, Manuscripts of the Greek 
Bible: An Introduction to Palaeography (New York, 1981), p. 99. 

72. The earliest known biblical manuscript that bears a date is appar- 
ently a palimpsest fragment of Isaiah in Syriac, written A.D. 459-60, now in 
the British Library; for a description of it, see E. Tisserant, "Le plus ancien 
manuscrit biblique date," Revue biblique, viii (1911), pp. 85-92. For an 
early dated Greek majuscule manuscript, see the description of Codex S. 

Important Witnesses to the Text of the New Testament 


Figure 15 Parchment leaf from a pocket-sized copy of the Book of Revela- 
tion (MS. 0169, fourth century), Speer Library, Princeton Theological Seminary, 
Princeton, New Jersey; Rev. 3-19-^.1. Actual size yA X 27» inches. 

library of Princeton Theological Seminary; see Fig. 15). Discovered at 
Oxyrhynchus in Egypt and dating from the fourth century, the page 
measures only 3% by 2"A inches — truly a pocket edition! 73 

73. One of the smallest Latin codices of the Gospels, measuring 5 by 
4 inches and furnished with illuminations, is described by Francoise Henry, 
"An Irish Manuscript in the British Museum (Add. 40618)," Journal of the 
Royal Society of Antiquaries of Ireland, lxxxvii (1957), pp. 147-66. Eight other 
examples of small books are discussed by Patrick McGurk, "The Irish Pocket 
Gospel Book," Sacris Erudiri, viii (1956), pp. 249-70. These manuscripts, all 
of which date from the seventh to ninth centuries, are written in a very small 
script, with a large number of abbreviations. They are also distinguished by a 
capricious and irregular use of one, two, or three columns to the page for the 
same manuscript and by all sorts of fanciful arrangements of the text on the 
page. For diminutive Greek and Coptic manuscripts, see W. H. Willis in Clas- 
sical, Mediaeval and Renaissance Studies in Honor of Bertbold Louis Ullman, 
ed. by Charles Henderson, Jr., i (Rome, 1964), p. 270, n. 1. C. H. Roberts 
discusses a variety of other miniature codices (a few of papyrus, most of 
parchment), the great majority of which contain Christian texts (Manuscript, 
Society and Belief in Early Christian Egypt [London, 1979], pp. 10-12). 

94 Thk Tkxt of thk New Tkstament 

What may well be the smallest codex (it is a biography of Mani in 
Greek, probably translated from a Syriac original) is about the size of 
a matchbox, each of the 192 pages measuring 1% by 1% inches. 74 The 
accepted criterion for description as a miniature is a breadth of less 
than 10 centimeters (4 inches). 

What is without doubt the largest biblical codex is the so-called 
Codex Gigas (the "giant codex"), now at Stockholm, which when lying 
open measures about 40 inches across the two pages and 36 inches 
high (see Fig. 17, p. 104). It is said that the hides of 160 asses were 
required for its production. 

II. Ancient Versions of the New Testament 75 

The earliest versions of the New Testament were prepared by 
missionaries, to assist in the propagation of the Christian faith among 
peoples whose native tongue was Syriac, Latin, or Coptic. Besides 

74. A preliminary report of the Mani Codex was made by A. Henrichs 
and L. Koenen, "Ein griechischer Mani-Codex (P. Colon. Inv. Nr. 4780),"' 
Zeitscbrift fur Papyrologie unci Epigrapbik, v (1970), pp. 97-216; edited, 
with a commentary, by the same authors, Zeitscbrift fur Papyrologie unci 
Epigrapbik, xix (1975), pp. 1-85 (pp. 1-72.7 of the codex), with the re- 
maining portions in sucessive volumes. Pages 1-99.8 of the codex are 
reproduced, with English translation by R. Cameron and A. J. Dewey, The 
Cologne Mani Codex (Missoula, MT, 1979); critical ed. with German trans, 
by L. Koenen and C. Romer (Opladen, 1988). 

For a discussion of the difference between an amulet and a miniature 
codex, see Michael J. Kruger, "P. Oxy. 840: Amulet or Miniature Codex?" 
Journal of 'Theological Studies, n.s., liii (2002), pp. 81-91- See also E. A. 
Judge, "The Magical Use of Scripture in (he Papyri," in Perspectives on 
Language and Text: Essays and Poems in Honor of Francis I. A ndersen 's 
Sixtieth Birthday, ed. by Edgar W. Conrad and Edward G. Newing 
(Winona Lake, IN, 1987), pp. 339-49, and S. R. Pickering, "The Significance 
of Non-continuous New Testament Textual Materials on Papyri," in Studies 
in the Early Text of the Gospels and Acts, ed. by D. G. K. Taylor ( Ihe Papers 
of the First Birmingham Conference on the Textual Criticism of the New 
Testament; Atlanta, 1999), pp. 121-40. 

75. For the history of scholarly research on the early versions, refer- 
ence may be made to Arthur Voobus, Early Versions of the New Testament, 
Manuscript Studies (Stockholm, 1954), and, for a full account of all versions 
of the New Testament prior lo A.D. 1000, see B. M. Metzger, 'Ihe Early 
Versions of the New Testament; '/heir Origin, Transmission, and Limitations 

Important Witnesses to the Text of the New Testament 95 

being of great value to the biblical exegete for tracing the history of 
the interpretation of the Scriptures, these versions are of no less im- 
portance to the textual critic in view of their origin in the second and 
third centuries. At the same time, however, it must be observed that 
there are certain limitations in the use of versions for the textual criti- 
cism of the New Testament. Not only were some of the translations 
prepared by persons who had an imperfect command of Greek, 76 but 
certain features of Greek syntax and vocabulary cannot be conveyed 
in a translation. For example, Latin has no definite article; Syriac can- 
not distinguish between the Greek aorist and perfect tenses; Coptic 
lacks the passive voice and must use a circumlocution. In some cases, 
therefore, the testimony of these versions is ambiguous. As for other 
questions, however, such as whether or not a given phrase or sen- 
tence was present in the Greek exemplar from which the translation 
was made, the evidence of the versions is clear and valuable. 77 

The study of the early versions of the New Testament is compli- 
cated by the circumstance that various persons made various transla- 
tions from various Greek manuscripts. Furthermore, copies of a trans- 
lation in a certain language were sometimes corrected one against 
the other or against Greek manuscripts other than the ones from 
which the translation was originally made. Thus, the reconstruction 
of a critical edition of an ancient version is often more complicated 
than the editing of the original Greek text itself. In tracing the inter- 
nal history of a version, however, the scholar has the advantage of 
divergent renderings (Ubersetzungsfarbe) for aid. Greek text types 
can be differentiated by variant readings alone, whereas in manu- 
scripts of the versions the same Greek reading may be represented by 

(Oxford, 1977). See also a thematic issue of Melanges de Science Religieuse, 
lvi, part 3 (1999), pp. 27-93, with articles on different versions of Mark 
(Latin, Coptic, Georgian, Palestinian Aramaic, and Arabic). 

76. Cf. St. Augustine's complaint of early translators of the Bible into 
Latin, that "no sooner did anyone gain possession of a Greek manuscript, 
and imagine himself to have any facility in both languages (however slight 
that might be), than he made bold to translate it" {De doctrina Christiana 
Il.xi [161). 

77. See A. F. J. Klijn, "The Value of the Versions for the Textual Criti- 
cism of the New Testament," in Jhe Bible Translator, trans, by H. H. Oliver, 
viii (1957), pp. 127-30. See also Allen Wikgren, "The Use of the Versions in 
New Testament Textual Criticism," Journal of Biblical Literature, lxvii 
(1948), pp. 135-41. 

96 The Text of the New Testament 

different renderings. By means of such variant renderings as well as 
variant readings, the several stages in the evolution of a version may 
be traced in the manuscript tradition. 

The most significant of the early versions of the New Testament 
are the following. 

1 . The Syriac Versions 78 

Scholars have distinguished five different Syriac versions of all or 
part of the New Testament. They are the Old Syriac, the Peshitta (or 
common version), the Philoxenian, the Harclean, and the Palestinian 
Syriac versions. That five or (possibly) six separate versions in Syriac 
were produced during the first six centuries of the Christian era is 
noteworthy testimony to the vitality and energy of Syrian church 
scholars. In fact, as Eberhard Nestle has reminded us, "No branch of 
the Early Church has done more than the Syriac-speaking. In our 
European libraries we have Syriac Bible MSS from Lebanon, Egypt, 
Sinai, Mesopotamia, Armenia, India (Malabar), even from China." 79 

a. The Old Syriac Version 

The Old Syriac Version of the four Gospels is preserved today in 
two manuscripts, both of which have large gaps. They are (1) a 
parchment manuscript now in the British Library, written in a clear 
and beautiful Estrangela hand (Fig. 16), edited by William Cureton in 
1858, and usually referred to as Syr c , and (2) a palimpsest manuscript 
that Agnes Smith Lewis discovered in the monastery of St. Catherine 
on Mount Sinai in 1892, called Syr\ Though these manuscripts were 
copied in about the fifth and fourth centuries, respectively, the form 
of text that they preserve dates from the close of the second or 
beginning of the third century. When the two manuscripts are com- 
pared, it is seen that the Sinaitic Syriac represents a slightly earlier 

78. Books useful in the study of the Syriac versions include the 
following: W. H. P. Hatch, An Album of Dated Syriac Manuscripts (Boston, 
1946), reprinted with a new foreword by Lucas Van Rompay (Piscataway, 
NJ, 2002); Terry C. Falla, A Key to the Peshitta Gospels, vol. 1 (Leiden, 
1991), vol. 2 (Leiden, 2000); George A. Kiraz, Comparative Edition of the 
Syriac Gospels: Aligning the Sinaiticus, Curetonianus, Peshitta, and 
Harklean Versions, 4 vols. (Leiden, 1996). 

79- "Syriac Versions," in Hastings Dictionary of the Bible, iv (1902), 
p. 645. 

Important Witnesses to the Text of the New Testament 




Figure 16 Curetonian Syriac MS. (early fifth century), British Library, 
London; John 6.30-41 (with headline and corrections). Actual size 11 K X 9 

foim of text than does the Curetonian, even though in some places 
it may have corruptions that the Curetonian has escaped. How far 
the text of the separated Gospels was influenced by the Gospel har- 
mony that Tatian prepared about A.D. 170 (see p. 134) has been 
much debated. In general, the Old Syriac version is a representative 
of the Western type of text. 

The Old Syriac version of Acts and the Pauline Epistles has 
not survived in extenso; we know it only through citations made by 
Eastern fathers. In the case of Acts, F. C. Conybeare reconstructed 
Ephraem's commentary from Armenian sources, of which a Latin 
translation is published in J, H. Ropes' The Text of Acts (London, 

98 The Text of the New Testament 

1926, pp. 373-453)- Ephraem's text of the Pauline Epistles was re- 
constructed by Joseph Molitor in Der Paulustext des hi. Kphrdm 
(Monumenta biblica et ecclesiastica, vol. iv; Rome, 1938). 

Editions: William Cureton, Remains of a Very Antient Recension of the 
Four Gospels in Syriac . . . (London, 1858); F. Crawford Burkitt, Hvangelion 
da-Mepharreshe: The Curetonian Version of the Four Gospels, with the 
Readings of the Sinai-Palimpsest . . . , 2 vols. (Cambridge, 1904); Agnes 
Smith Lewis, The Old Syriac Gospels . . . (London, 1910); Arthur Hjelt, Syrus 
Sinaiticus (Helsingfors, 1930) Iphotographic facsimile). 

b. The Peshitta Version 

The Peshitta version, or Syriac Vulgate, of the New Testament 
(Syr p ) was prepared about the beginning of the fifth century, proba- 
bly in order to supplant the divergent, competing Old Syriac transla- 
tions. It contains only 22 books; 2 Peter, 2 and 3 John, Jude, and 
Revelation were not translated. Until recendy, scholars thought that 
Rabbula, Bishop of Edessa (c. A.D. 411-31), was responsible for the 
Peshitta; but it is more likely that his revision marked an intermedi- 
ate stage between the Old Syriac text and the final form of the 
Peshitta. 80 Because the Peshitta was accepted as the standard version 
of the Scriptures by both Eastern and Western branches of Syrian 
Christendom, one must conclude that it had attained some degree of 
status prior to the split in the Syrian Church in A.D. 431. 

More than 350 manuscripts of the Peshitta New Testament are 
known today, several of which date from the fifth and sixth cen- 
turies. The text of the Peshitta has been transmitted with remarkable 
fidelity, so that very few significant variants exist among the wit- 
nesses. The textual complexion of the Peshitta version has not yet 
been satisfactorily investigated, but apparently it represents the work 
of several hands in various sections. In the Gospels, it is closer to the 
Byzantine type of text than in Acts, where it presents many striking 
agreements with the Western text. 81 

80. See Arthur Voobus, Studies in the History of the Gospel Text in Syriac 
(Louvain, 1951). 

81. For an account of multiple frustrations over the years that have 
prevented the production of a critical edition of the Peshitta Acts and 
Epistles, based on newly made collations, see Roderick Grierson, "'Without 
Note or Comment': British Library Or 1 1350 and the Text of the Peshitta 
New Testament," Oriens Christianus, lxxxii (1998), pp. 88-98. 

Important Witnesses to the Text of the New Testament 99 

Editions: P. E. Pusey and G. H. Gwilliam, Tetraevangelium sanctum 
iuxta simplicem Syrorum versionem adfidem codicum . . . (Oxford, 1901) 
[based on 42 manuscripts; with a critical apparatus and a Latin translation]; 
The New Testament in Syriac (London, 1905-20) [published by the British 
and Foreign Bible Society; the Gospels are reprinted from Pusey and 
Gwilliam's text (without apparatus), and the rest of the New Testament was 
edited by Gwilliam and J. Gwynn]; and Das Neue Testament in Syrischer 
(Jberlieferung, I, Die grossen katholischen Briefe, ed. by Barbara Aland 
(Berlin, 1986); II. Die Paulinischen Briefe, ed. by Barbara Aland and Andreas 
Juckel, Teil 1 (Berlin, 1991), Teil 2 (Berlin, 1995), Teil 3 (Berlin, 2002). 

c. The Philoxenian and/or Harclean Version(s) 

One of the most confused and confusing tangles of textual criti- 
cism involves the unraveling of the Philoxenian and/or Harclean ver- 
sion^), usually abbreviated Syr ph and Syr 11 , respectively. The scanty 
evidence in several colophons found in certain Harclean manuscripts 
has been interpreted in quite different ways. On the one hand, it has 
been held that the Syriac version produced in A.D. 508 for Philox- 
enus, Bishop of Mabbug, by Polycarp his chorepiscopus was reis- 
sued in 616 by Thomas of Harkel (Heraclea), Bishop of Mabbug, 
who merely added marginal notes derived from two or three Greek 
manuscripts. On the other hand, it has been held that the Philoxen- 
ian version was thoroughly revised by Thomas, who also added in 
the margin certain readings that he considered to be important but 
not worthy of inclusion in the text. In other words, according to the 
first view, there is but one version that was republished with variant 
readings noted in the margin; according to the second, there are two 
separate versions entirely, the later one being provided with margin- 
alia. It is not necessary to attempt to resolve this complicated prob- 
lem here; in any case, during the sixth century, for the first time in 
the history of the Syriac-speaking churches, the minor Catholic Epis- 
tles and Revelation were translated into Syriac. The Harclean appa- 
ratus of Acts is the second most important witness to the Western 
text, being surpassed in this respect only by Codex Bezae. 

Editions: Joseph White, Sacrorum Evangeliorum Versio Syriaca Philox- 
eniana (Oxford, 1778); idem, Actuum Apostolorum et Epistolarum tarn 
Catholicarum quam Paulinarum Versio Syriaca Philoxeniana (Oxford, 
1799-1803); R- L- Bensly, The Harklean Version of the Epistle to the Hebrews, 
Chap. XI. 28-XIII. 25 (Cambridge, 1889); John Gwynn, The Apocalypse of 
St. John, in a Syriac Version Hitherto Unknown . . . (Dublin and London, 
1897); idem, Remnants of the Later Syriac Versions of the Bible . . . The Four 

100 The Text of the New Testament 

Minor Catholic Epistles in the Original Philoxenian Version . . . and John VII. 
52- VIII. 12 . . . (London and Oxford, 1909). iThe text of the Apocalypse 
and of the minor Catholic Epistles is included in the British and Foreign 
Bible Society's edition of the Peshitta.] 

d. The Palestinian Syriac Version 

The translation into Christian Palestinian Syriac (i.e., Aramaic) is 
known chiefly from a lectionary of the Gospels, preserved in three 
manuscripts dating from the eleventh and twelfth centuries. In addi- 
tion, fragments of the Gospels, in a continuous text, are extant, as are 
scraps of Acts and of several of the Pauline Epistles. When this ver- 
sion (abbreviated Syr pal ) was made has been much disputed, but 
most scholars think that it dates from about the fifth century. Appar- 
ently, it is based on a Greek text of the Caesarean type and is quite 
independent of the other Syriac versions. 

Editions: Agnes Smith Lewis and Margaret Dunlop Gibson, The Pales- 
tinian Syriac Lectionary of the Gospels (London, 1899); Agnes Smith Lewis, 
Codex Climaci rescriptus (Horae semiticae, viii, Cambridge, 1908) [contains 
fragments of the Gospels, Acts, and Pauline Epistles]; for other fragments, 
see Cyril Moss, Catalogue of Syriac Printed Books and Related Literature in 
the British Museum (London, 1962), to which should be added Nina V. 
Pigulevskaya, Katalog siriiskikh rukopisei Leningrada iPalestinskii sbornik, 
vi [691, Moscow, I960, pp. 3-230). 82 

2. The Latin Versions 83 

When and where it was that the earliest attempts were made to 
translate the Bible into Latin has been much disputed. In the opinion 
of most scholars today, the Gospels were first rendered into Latin 

82. For a convenient scriptural list indexing the several publications 
containing portions of the Palestinian Syriac version, see Fr. Schulthess, 
lexicon Syropalaestinum (Berlin, 1903), pp. vii-xvi. For additions to the list 
in the Book of Acts, see Charles Perrot, "Un fragment christo-palestinien 
decouvert a Khirbet Mird (Actes des Apotres, x, 28-29; 32-41)," Revue 
Biblique, Ixx (1963), pp. 506-55, especially 544. For Syriac manuscripts in 
general, see James T. demons, "A Checklist of Syriac Manuscripts in the 
United States and Canada," Orientalia Christiana Periodica, xxxii (1966), 
pp. 224-51, 478-522. 

83- See J. K. Elliott, "The Translations of the New Testament into Latin: 
The Old Latin and the Vulgate," in Aufstieg und Niedergang der Rbmischen 

Important Witnesses to the Text of the New Testament 101 

during the last quarter of the second century in North Africa, where 
Carthage had become enamored of Roman culture. Not long after- 
ward, translations were also made in Italy, Gaul, and elsewhere. The 
wooden and literalistic style that characterizes many of these render- 
ings suggests that early copies were made in the form of interlinear 
renderings of the Greek (see Fig. 13, p. 82). 

a. The Old Latin Version(s) 

During the third century, many Old Latin versions circulated in 
North Africa and Europe, including distinctive versions that were 
current in Italy, Gaul, and Spain. Divergent renderings of the same 
passage (e.g., at Luke 24.4-5 there are at least 27 variant readings in 
the Old Latin manuscripts that have survived) bear out Jerome's 
complaint to Pope Damasus that there were almost as many versions 
as manuscripts (.tot enim sunt exemplaria paene quot codices)^ 

No codex of the entire Old Latin Bible is extant. The Gospels are 
represented by about 32 mutilated manuscripts, besides a number of 
fragments. About a dozen manuscripts of Acts are extant. There are 
four manuscripts and several fragments of the Pauline Epistles but 
only one complete manuscript and several fragments of the Apoca- 
lypse. These witnesses date from the fourth century to the thirteenth 
century, thus proving that the Old Latin version was still copied long 
after it had gone out of general use. The Old Latin manuscripts are des- 
ignated in a critical apparatus by small letters of the Latin alphabet. 85 

The textual complexion of the Old Latin version(s) is typically 
Western. As a rule, the form of the Old Latin current in Africa pre- 
sents the greater divergences from the Greek and the form(s) current 
in Europe, the smaller. 

The most important witnesses of the Old Latin versions are the 
following (grouped according to the African and European types 
of text). 

Welt, II. 26.1 (Berlin, 1992), pp. 198-245; Jacobus H. Petzer, "The Latin Ver- 
sion of the New Testament," in The New Testament in Contemporary Re- 
search, ed. by Ehrman and Holmes, pp. 1 13-30; Philip Burton, The Old 
Latin Gospels: A Study of Their Texts and Language (Oxford, 2000). 

84. See St. Jerome's preface (Novum opus) to his translation of the 
four Gospels. 

85. For a register of Old Latin biblical manuscripts, each with a full 
description and select bibliography, see Roger Gryson, Altlateinische 
Handschriften. Manuscrits vieux latins, 1 (Freiburg, 1999). 

102 The Text of the New Testament 


e. Codex Palatinus, designated by the symbol e, is a fifth-century 
manuscript containing portions of the four Gospels, written with sil- 
ver ink on purple parchment. Though the type of text in e is basically 
African, it has been strongly Europeanized. Augustine probably em- 
ployed a Gospel text of this kind before A.D. 400. 

h. The symbol h is given to the fragmentary sixth-century man- 
uscript known as the Fleury palimpsest, which contains about one- 
quarter of Acts besides portions of the Catholic Epistles and the Book 
of Revelation. The manuscript presents many scribal errors, and the 
rendering into Latin is often very free; for example, the narrative of 
Paul's voyage, Acts 28.1-13, appears to be a corrupt form of an 
abridgement made by the translator. 

k. The most important witness to the African Old Latin is Codex 
Bobbiensis, to which the symbol k has been assigned. Unfortunately, 
it is quite fragmentary, containing only about half of Matthew and 
Mark. It was copied about A.D. 400 in Africa and brought to the Irish 
monastery of Bobbio in northern Italy, where it was preserved for 
many centuries until it found a home in the National Library at Turin, 
where it is now, Its form of text agrees very closely with the quotations 
made by St. Cyprian of Carthage (about A.D. 250). According to E. A. 
Lowe, 86 k shows paleographical marks of having been copied from a 
second-century papyrus. It is noteworthy that k contains the interme- 
diate ending of the Gospel according to Mark (see § 2, p. 323)- 


a. What is probably the oldest European manuscript of the 
Gospels is Codex Vercellensis (known by the symbol a), kept in the 
cathedral treasure room at Vercelli in northern Italy. According to an 
old tradition, it was written by the hand of St. Eusebius, Bishop of 
Vercelli, who was martyred in 370 or 371. Next to k, it is the most 
important Old Latin manuscript of the Gospels. 

b. Codex Veronensis (b), in the possession of the Chapter Library 
of the cathedral at Verona, Italy, is a purple parchment manuscript 
written in the fifth century with silver and occasionally gold ink. It 
contains the four Gospels (almost in their entirety) in the order of 
Matthew, John, Luke, and Mark. In Burkitt's opinion, it represents the 
type of text that Jerome used as the basis of the Vulgate. 

86. Reported by D. Plooij in Bulletin of 'the Bezan Club, xi (1936), p. 11. 

Important Witnesses to the Text of the New Testament 103 

c. Codex Colbertinus, written in the twelfth century, probably in 
southern France, is now in the Bibliotheque Nationale at Paris and 
contains the four Gospels in a mixed form of text. Clear traces of 
African readings persist in what is generally a European Old Latin 
text contaminated here and there by Jerome's Vulgate. 

d. The Latin side of the fifth-century bilingual Codex Bezae (D), 
though corrected here and there from the Greek side, preserves an 
ancient form of Old Latin text. Since d agrees occasionally with read- 
ings of k and of a when all other authorities differ, it witnesses to a 
text that was current no later than the first half of the third century 
and may be earlier still. 

ff 2 . Codex Corbiensis is a mutilated copy of the four Gospels, 
of the fifth or sixth century, formerly belonging to the monastery 
of Corbey, near Amiens, and now in the Bibliotheque Nationale at 
Paris. It contains a form of text akin to that preserved in a and b. 

gig. What is undoubtedly one of the largest manuscripts in the 
world is appropriately named Codex Gigas (the "giant"). 87 Each page 
measures about 20 by 36 inches, and when the codex lies open it 
makes an impressive sight. Written in the early part of the thirteenth 
century at the Benedictine monastery of Podla^ice in Bohemia, it 
was later acquired by the Imperial Treasury in Prague. When the 
Swedish army conquered the city in 1648, it was brought to Sweden 
and presented to the Royal Library in Stockholm the following year. 

In addition to the text of the entire Bible in Latin, the "giant" 
manuscript contains Isidore of Seville's Htymologiae (a general ency- 
clopedia in 20 books), a Latin translation of Flavius Josephus' Antiq- 
uities of the Jews, Cosmas of Prague's Chronicle of Bohemia, as well 
as other works. It is sometimes called the Devil's Bible (Djcivuls- 
bibelri) because fol. 290 contains a huge picture of that potentate in 
garish colors, with horns, forked tongue, and claws on fingers and 
toes (Fig. 17). According to legend, the scribe was a monk who had 
been confined to his cell for some breach of monastic discipline and 
who, by way of penance, finished the manuscript in a single night 
with the aid of the devil, whom he had summoned to help him. 

Codex Gigas is of importance to textual critics because the Book 
of Acts and the Book of Revelation preserve a form of Old Latin text 

87. For a description of the manuscript and its contents, see B. Dudik, 
Forschungen in Schweden fur Mcihrens Geschichte (Briinn, 1852), 
pp. 207-35. 


The Text of the New Testament 

Figure 17 Codex Gigas (A.D. 1204—30), Royal Library, Stockholm, some- 
times called the Djdvnlsbibeln (Devil's Bible) because of this picture on folio 
i90 (see pp. 94 and 103). Actual size 36 X 20 inches. 

that agrees with the scriptural quotations made by Bishop Lucifer of 
Cagliari (in Sardinia) about the middle of the fourth century. 

m. The symbol m is used to refer to a patristic collection of bib- 
lical passages arranged by topics to illustrate special points f con- 
duct. The treatise is frequently called Speculum (the Latir word 
meaning "mirror" [for conduct]) and is preserved in a number of 
manuscripts, of which the oldest is of the eighth or ninth century. 
The scriptural quotations are in a Spanish form of the African Old 
Latin text, agreeing (in the Catholic Epistles) almost ad litteram with 

Important Witnesses to the Text of the New Testament 105 

the quotations of Priscillian, who in 385 was condemned at Treves 
for magic (maleficium) and executed, the first person to be put to 
death by the Church. 

Editions: The more important manuscripts of the Old Latin Bible have 
been published in two series entitled Old Latin Biblical Texts (Oxford, 1883 
onward) and Collectanea biblica latina (Rome, 1912 onward). The most sat- 
isfactory edition of the Old Latin texts of the Gospels is the series entitled 
Itala: das Neue Testament in altlateinischer Uberlieferung, begun by Adolf 
Jiilicher and continued by W. Matzkow and Kurt Aland. The Vetus-Latina- 
Institut at the monastery of Beuron in Wurttemberg, Germany, under the 
leadership of the late Fr. Bonifatius Fischer, has issued fascicles of an edition 
entitled Vetus Latina: die Keste der altlateinischen Bibel. This ambitious pro- 
ject seeks to assemble from manuscripts and quotations made by Church fa- 
thers all the evidence of the Latin Bible as it circulated prior to the revision 
undertaken by St. Jerome. 

b. The Latin Vulgate. Toward the close of the fourth century, the 
limitations and imperfections of the Old Latin versions became evident 
to leaders of the Roman Church. It is not surprising that about A.D. 382 
Pope Damasus requested the most capable biblical scholar then living, 
Sophronius Eusebius Hieronymus, known today as St. Jerome, to un- 
dertake a revision of the Latin Bible. Within a year or so, Jerome was 
able to present Damasus with the first fruits of his work, a revision of 
the text of the four Gospels, where the variations had been extreme. 
In a covering letter, he explained the principles that he followed: he 
used a relatively good Latin text as the basis for his revision and 
compared it with some old Greek manuscripts. He emphasized that 
he treated the current Latin text as conservatively as possible and 
changed it only where the meaning was distorted. Though we do not 
have the Latin manuscripts that Jerome chose as the basis of his work, 
it appears that they belonged to the European form of the Old Latin 
(perhaps they were similar to manuscript b). The Greek manuscripts 
apparently belonged to the Alexandrian type of text. 

When and how thoroughly Jerome revised the rest of the New 
Testament has been much debated. Several scholars (De Bruyne, 
Cavallera, B. Fischer) have argued that Jerome had nothing to do 
with the making of the Vulgate text of the rest of the New Testament 
but that, by a curious twist of literary history, the work of some other 
translator came to be circulated as Jerome's work. The commonly ac- 
cepted view, however, rests upon the natural interpretation of what 
Jerome says about his work of revision. In either case, it is apparent 

106 The Text of the New Testament 

that the rest of the New Testament was revised in a much more cur- 
sory manner than were the Gospels. 88 

It was inevitable that, in the transmission of the text of Jerome's 
revision, scribes would corrupt his original work, sometimes by care- 
less transcription and sometimes by deliberate conflation with copies 
of the Old Latin versions. In order to purify Jerome's text, a number 
of recensions or editions were produced during the Middle Ages; no- 
table among these were the successive efforts of Alcuin, Theodulf, 
Lanfranc, and Stephen Harding. Unfortunately, however, each of 
these attempts to restore Jerome's original version resulted eventu- 
ally in still further textual corruption through mixture of the several 
types of Vulgate text that had come to be associated with various 
European centers of scholarship. As a result, the more than 8,000 
Vulgate manuscripts that are extant today exhibit the greatest amount 
of cross-contamination of textual types. 

The most noteworthy of these manuscripts include the following 
(they are usually denoted by capital letters or sometimes by the first 
syllable of their names). 

A. Codex Amiatinus, dating from the seventh or eighth century 
in the Laurentian Library at Florence, is a magnificent manuscript 
containing the whole Bible and weighing 75 pounds. It was written 
by order of Ceolfrid, abbot of Jarrow and Wearmouth, and sent by 
him as a gift to Pope Gregory in 716. Many scholars regard it as the 
best manuscript of the Vulgate (Fig. 18). 

C. Codex Cavensis, dating from the ninth century, is in the La 
Cava monastery near Salerno. It contains the whole Bible and is one 
of the chief representatives of the Spanish group of manuscripts. 

D. Codex Dublinensis, or the Book of Armagh, is at Trinity 
College, Dublin. Dating from the eighth or ninth century, it con- 
tains the whole New Testament as well as the apocryphal Epistle 
of Paul to the Laodiceans. It presents the Irish type of Vulgate text, 
which is characterized by small additions and insertions. Here and 
there it shows signs of having been corrected from Greek manu- 
scripts akin to the Ferrar group (fam. 13). In the Pauline Epistles, 
the text is predominately Old Latin, with only a minority of Vulgate 

88. For a recent survey of issues involved with Jerome's translation, 
see Catherine Brown Tkacz, "Labor tarn utilis: The Creation of the 
Vulgate," Vigiliae Christianae 1 (1996), pp. 42-72. 

Important Witnesses to the Text of the New Testament 107 


Bdra [>o FWVIN* Hoi Mr.\a.\vir ofvs 


Figure 18 Codex Amiatinus (seventh or eighth cent y), Laurentian 
Library, Florence; Ezra the Scribe (see pp. 27 and 106). 

108 The Text of the New Testament 

F. Codex Fuldensis, now in the Landesbibliothek at Fulda, was 
written between A.D. 541 and 546 at Capua by order of Victor, the 
bishop of that see, and was corrected by him personally. It contains 
the whole New Testament together with the apocryphal Epistle to 
the Laodiceans. The Gospels are arranged in a single, consecutive 
narrative, in imitation of Tatian's diatessaron (see pp. 131-3)- Its text, 
which is very good, is akin to that of Codex Amiatinus. 

M. Codex Mediolanensis, in the Ambrosian Library at Milan, is a 
Gospel manuscript of the early sixth century. In the judgment of 
Wordsworth and White, it ranks with Amiatinus and Fuldensis as one 
of the best witnesses of the Vulgate. 

R. The Rushworth Gospels, now in the Bodleian Library, is also 
known as "the Gospels of Mac-Regol," from the name of the scribe, 
who died in A.D. 820. It contains a Latin Vulgate text marked by a 
number of Old Latin readings, fairly common to Irish manuscripts, In 
the latter half of the tenth century, it was provided with an interlinear 
Anglo-Saxon gloss (see p. 125). 

Y. The celebrated Lindisfarne Gospels, of about A.D. 700, now in 
the British Library, is a beautifully executed codex, adorned with 
Celtic-Saxon illumination. It is furnished with an Anglo-Saxon inter- 
linear gloss, the earliest form of the Gospels in the ancestor of 
English. Its Latin text is closely akin to that of Codex Amiatinus. 

2. Codex Harleianus, formerly in the Royal Library at Paris, was 
stolen from there, as it seems, by Jean Aymon in 1707 and sold to 
Robert Harley, who deposited it in the British Museum. It is a beauti- 
fully written copy of the Gospels, dating from the sixth or seventh 

Z. Codex Sangallensis, the oldest known manuscript of the 
Vulgate Gospels, was written in Italy probably toward the close of 
the fifth century. 89 More than half of it survives at the monastery of 
St. Gall and in other libraries. Unfortunately, Wordsworth and White 
overlooked this important manuscript in preparing their edition of 
the Vulgate. 

p. What has been called one of the finest, if not the finest, of 
purple manuscripts in existence is the Golden Gospels, now in the 
J. Pierpont Morgan Library, New York. Written entirely in letters of 

89. Lowe thought that it may have been copied "possibly during the 
lifetime of Jerome"; see E. A. Lowe, Codices Latini Antiquiores, vii (Oxford, 
1956), p. 41. 

Important Witnesses to the Text of the New Testament 109 

burnished gold on purple parchment, this sumptuous codex con- 
tains a Vulgate Latin text with Northumbrian and Irish affinities. Pre- 
viously thought to date from the close of the seventh or the begin- 
ning of the eighth century, it has lately been assigned to the tenth 
century. 90 

Editions of the Vulgate: The decision of the Council of Trent (1546) to 
prepare an authentic edition of the Latin Scriptures was finally taken up by 
Pope Sixtus V, who authorized its publication in 1590. The Sixtine Vulgate 
was issued with a papal bull threatening major excommunication for viola- 
tors of the commands that variant readings should not be printed in subse- 
quent editions and that the edition must not be modified. 91 (According to 
Steinmuller, 92 however, this bull "today is commonly recognized as not hav- 
ing been properly and canonically promulgated.") In 1592, after the death of 
Sixtus, Pope Clement VIII called in all the copies he could find and issued 
another authentic edition — differing from the former in some 4,900 variants! 
This latter edition remains the official Latin Bible text of the Roman Catholic 
Church to the present day. 

A critical edition of the New Testament, with an apparatus, was pub- 
lished at Oxford by a group of Anglican scholars. Begun by Bishop John 
Wordsworth and H. J. White, the first volume, containing the text of the four 
Gospels, was issued in 1899; the last volume, containing the Book of Reve- 
lation, was completed by H. F. D. Sparks in 1954. For a judicious appraisal 
of the adequacy of the Oxford Vulgate (the volume containing the Gospels 
suffered particularly from the editors' inexperience), see Bonifatius Fischer, 
Zeitschrift fur die neutestamentliche Wissenschqft, xlvi (1955), pp. 178-96. 

In 1908, Pope Pius X appointed the International Commission for Revi- 
sion of the Vulgate. Publication began with Genesis at the Vatican Press 
in 1926. A handy two-volume edition with a brief apparatus prepared by 
R. Weber, O.S.B., and four collaborators was published by the Wiirtemberg 
Bibelanstalt at Stuttgart in 1969 (4th ed. by R. Gryson, 1994). 

90. The earlier date was advocated by Wattenbach, de Rossi, Gregory, 
and its editor, H. C. Hoskier, The Golden Latin Gospels in the Library of 

]. Pierpont Morgan (New York, 1910); the later date was subsequently pro- 
posed by F. A. Lowe, "The Morgan Golden Gospels: The Date and Origin 
of the Manuscript," in Studies in Art and Literature for Belle da Costa 
Greene, ed. by Dorothy Miner (Princeton, 1954), pp. 266-79. 

91. On the Sixtine edition of the Vulgate Bible, see Paul M. Baumgarten, 
Die Vulgata Sixtina von 1590 und ihre Einfuhrungsbulle, Aktenstilcke und 
Untersuchungen (Munster, 1911). 

92. John E. Steinmuller, A Companion to Scripture Studies, i (New York, 
1941), p. 192. 

110 The Text oh the New Testament 

3. The Coptic Versions 93 

Coptic is the latest form of the ancient Egyptian language, which 
until Christian times was written in hieroglyphs and their two deriv- 
atives, hieratic and demotic script. In the first centuries of the Chris- 
tian era, the language came to be written in Greek majuscules, with 
the addition of seven characters taken from the demotic. 

During the early Christian period, the old Egyptian language was 
represented in at least half a dozen dialectal forms throughout Egypt, 
differing from one another chiefly in phonetics but also to some 
extent in vocabulary and syntax. In the southern part of the country, 
called Upper Egypt, the Sahidic dialect prevailed from Thebes to the 
south. Around the delta in the northern part of Egypt, called Lower 
Egypt, the Bohairic dialect was used along with Greek. At various 
settlements along the Nile between these two parts of the country, 
there developed intermediate dialects, chiefly the Fayyumic (formerly 
known as Bashmuric), Memphitic (or Middle Egyptian), Achmimic, 
and sub-Achmimic (used south of Asyut). 

Of these dialects, the Sahidic and the Bohairic are the most 
important for the study of early versions of the Bible. About the 
beginning of the third century, portions of the New Testament were 
translated into Sahidic, and within the following century most of 
the books of the New Testament became available in that dialect. 
Indeed, to judge on the basis of widely divergent Sahidic texts, some 
parts of the Scriptures were translated at various times by independent 
translators. In general, the Sahidic version agrees with the Alexandrian 
form of text, but in the Gospels and Acts it has many Western 
readings. Unfortunately, when Horner prepared his edition of this ver- 
sion, only fragmentary manuscripts were available. 94 Subsequently, 

93. See Gerd Mink, "Die Koptischen Versionen des Neuen Testaments,'' 
Die alien Ubersetzungen des Neuen Testaments, die Kircbenvdterzitate und 
Lectionare, ed. by K. Aland (Berlin, 1972), pp. 160-200; Frederik Wisse, "The 
Coptic Versions of the New Testament," in The Text of the New Testament in 
Contemporary Research, ed. by Ehrman and Holmes, pp. 131-41. 

94. What has been thought to be the oldest Sahidic manuscript of those 
used by Horner is a papyrus codex containing portions of Deuteronomy, 
Jonah, and the Acts of the Apostles (B. M. Or. 7594), ed. by E. A. W. Budge, 
Coptic Biblical Texts in the Dialed of Upper Egypt (London, 1912). Fol. 108 v 
contains a section written in Coptic but in a cursive Greek hand that Kenyon 
assigned (op. cit., Introduction, p. Ixiii) to about the middle of the fourth 

Important Witnesses to the Text of the New Testament 111 

;»M*l'c»Eo* t5l*rfc»ictJ«Atiaii»« 

^ j»M"|TB S»<*Yl-r»Me«|*MUtt»f P* 


•fS-ro**! IK V*AYuJ<U*rK>< t If* H*|Jk«| 

. puiuc y>»v*y>v>v>>>s-> 
, TASU(J'4t«|*'A>u*rf6|dtf)uYV«(' 

>»»*»•*>* e*FT5.««or<uri'»#' 

ete loKpi'iwn^'MeneiiiM** 

M**ruMPM«?f'UW'TF«** ,, rfc<r*nrt 

06IM* • KIVoriMU uaci'i ~ x * 
OPim'iwiK. hi«ihc-<:<|M#»> 
eFii<«»ct«iic-Mc*i; , *»»iK«»c 

Figure 19 Chester Beatty Coptic (Sahidic) MS. B (about A.D. 600), Chester 
Beatty Museum, Dublin, John 1.1-10. Actual size 4X X 4 inches. 

the Pierpont Morgan Library in New York acquired a large collection 
of Coptic manuscripts, most of which are complete. Among them is 
a Sahidic tetraevangelium (Morgan MS. 569) dating from the eighth 
or ninth century and preserving Matthew, Mark, and John in their 
entirety; Luke lacks 14 leaves. The collection of manuscripts is now 
available in photographic reproduction prepared by Henri Hyvernat. 
A. Chester Beatty also acquired three Sahidic manuscripts dating from 
about the sixth or seventh century. One of them (Codex D) contains 
the Acts of the Apostles, followed by the Gospel according to John 
(Fig. 19); another (Codex A) preserves the Pauline Epistles, followed 

century. Ad. Hebbelynck, however, raised questions about the unity and the 
age of the manuscript (Museon, xxxiv [1921], pp. 71-80), and II. Hyvernat 
thought a case could be made for dating it in the sixth century (quoted by 
Lagrange, Critique textuelle, ii [Paris, 1935], p. 324, n. 2). Sir Herbert 
Thompson, it should be added, accepted Kenyon's fourth-century dating 
(The Coptic Version of tbe Acts of the Apostles . . . [Cambridge, 1932], p. xxi). 

112 The Text or the New Testament 

by the fourth Gospel. The third codex is fragmentary and contains 
the text of Psalms 1-50 with the first chapter of Matthew. Thompson 
published an edition of the text of the Beatty Acts and Pauline Epis- 
tles, with a collation of the text of John against Horner's edition. 

The Bohairic version appears to be somewhat later than the 
Sahidic version. It survives in many manuscripts, almost all of them of 
a very late date (the earliest complete Gospel codex still extant was 
copied A.D. 1174). In 1958, an early papyrus codex containing most 
of the Gospel of John and the opening chapters of Genesis in Bohairic 
was published in the Bodmer series. 95 Although the first few folios are 
badly mutilated, beginning at about the middle of the fourth chapter 
of John the text is much better preserved. The editor, Rodolphe 
Kasser, was inclined to date the manuscript in the fourth century. It 
is of interest that passages that textual scholars have regarded as 
critically suspect (such as the statement about the angel moving the 
water in John 5.3b— 4 and the pericope de adultera, 7.53-8.11) are not 
present in this manuscript. The Greek prototype of the Bohairic ver- 
sion appears to be closely related to the Alexandrian text type. 

Among the scattered manuscripts that preserve portions of the 
New Testament in the Fayyumic dialect, one of the earliest is a pa- 
pyrus codex, now at the University of Michigan, which contains 
John 6.11-15.11 (with lacunae). According to its editor, Elinor M. 
Husselman, the manuscript dates from the early part of the fourth 
century. In textual affinities, it agrees roughly twice as often with 
the Sahidic version as it does with the Bohairic. 

The most significant representative of the sub-Achmimic version 
is a papyrus codex containing the Gospel of John. In the opinion of 
its editor, Sir Herbert Thompson, the manuscript dates from about 
A.D. 350-75. Like the Sahidic version, with which it is related, the 
version appears to be a representative of the Alexandrian type of text. 

Of the 80 manuscripts of the Gospels in Sahidic that have been 
catalogued by Schmitz and Mink, 96 three of the more important, con- 
taining Mark, Luke, and John and dating from the fifth century, have 

95. Another early manuscript of the New Testament in Bohairic (actu- 
ally it is semi-Bohairic) is a fourth- or early fifth-century parchment fragment 
of Philippians, edited by Paul E. Kahle in Museon, lxiii (1950), pp. 147-57. 

96. Liste der koptischen Handschriften des Neuen Testaments: J. Die 
sahidischen Handschriften der Evangelien, 1. Teil, ed. by F.-J. Schmitz and 
G. Mink (Arbeilen zur Neutestamentlichen Textforschung, viii; Berlin, 
1986); 2. Teil, 1. Halbband (Arbeiten . . . , xiii; Berlin, 1989). 

Important Witnesses to the Text of the New Testament 113 

been edited by Hans Quecke. 97 From a textual point of view, the 
three manuscripts offer more or less what might have been 
expected, often agreeing with witnesses of the Alexandrian type of 
text. Thus, Mark ends at 16.8 in company with B and «. In chapter 1, 
with the first hand of «, the manuscript more than once presents the 
shorter reading; for example, it lacks "Son of God" at verse 1, "and 
saying" at verse 15, and "that he was the Christ" at verse 34. In com- 
mon with other Sahidic witnesses as well as p 75 the name "Nineveh" 
is given to the rich man in Luke 16.19- Along with p 75 , the text of 
John 10.7 reads "shepherd" (instead of "door"), and at 21.6, along 
with p 66 and other witnesses, it contains the addition from Luke 5.5. 
At 11.12, it reads "will rise" (instead of "will recover") along with p 75 , 
the sub-Achmimic, and the Bohairic. 

In recent years, several copies of biblical manuscripts written in 
the Middle Egyptian dialect (also known as the Oxyrhynchite dialect) 
have come to light. Noteworthy among them are two papyrus 
codices of the same dimensions (5 by AV» inches), dating from the 
fourth or fifth century and containing, respectively, Matthew and 
the first half of Acts. The former, now in the possession of the 
Scheide Library of Princeton, New Jersey, is one of the four oldest 
copies of the entire text of Matthew. 98 Edited by Hans-Martin 
Schenke, 99 its textual affinities are closely related to Codex Vaticanus 
and Codex Sinaiticus in sometimes supporting the shorter reading. 
These include the absence of the doxology at the close of the Lord's 

97. Das Markusevangelium saidisch: Text der Handschrift PPalau Rib. 
Inv.-Nr. 182 mit der Varianten der Handschrift M569 (Barcelona, 1972); 
Das Liikasevangelium saidisch: Text der Handschrift PPalau Rib. lnv.-Nr. 
181 mit den Varianten der Handschrift M569 (Barcelona, 1977); Das 

fohannesevangelium saidisch: Text der Handschrift PPalau Rib. Inv.-Nr. 
183 mit den Varianten der Handschriften 813 und 814 der Chester Beatty 
Library und der Handschrift M569 (Barcelona, 1984). 

98. Of the three others, Codices Vaticanus and Sinaiticus belong to the 
mid-fourth century and Codex Washingtonianus is dated to the fifth cen- 
tury. Codices Alexandrinus and Kphraemi of the fifth century are incom- 
plete in Matthew, and the several Greek papyri that antedate the sixth cen- 
tury preserve only scraps of the text of Matthew. 

99. Das Matthaus-Evangelium im Mitteldgyptischen Dialekt des Koptis- 
chen {Codex Scheide), ed. by Hans-Martin Schenke (Texte und Unter- 
suchungen, cxxvii; Berlin, 1981), with facsimile reproduction of selected 
folios of the codex. 

114 The Text of the New Testament 

Prayer (Matt. 6.13), the signs of weather (16.2-3), the statement 
about the Son of Man (18.11), and one of the woes of the Pharisees 
(23.14). On the other hand, the manuscript not infrequently agrees 
with one or another group of witnesses in support of the longer 
reading. These include the presence of 12.47 as well as the ancient 
reading xai zfjg vv[i(piag (25.1), which is supported by a wide range 
of Western and versional witnesses (D, X*, fam. 1, Old Latin, Vulgate, 
Syriac s,p,h 7 Armenian, Georgian 16 , Diatessaron). 100 

The Acts manuscript in Middle Egyptian was purchased by the 
late William S. Glazier and is at present in the keeping of the Pierpont 
Morgan Library of New York City. Preliminary analyses of the text in- 
dicate that it is a notable representative of the so-called Western type 
of text. 101 

Portions of ten Pauline Epistles in Middle Egyptian are preserved 
in a fragmentary papyrus codex now at the Istituto di papirologia of 
the University of Milan. Dated by the editor 102 to the first half of the 
fifth century, the surviving fragments contain portions of the follow- 
ing (in this order): Romans, 1 and 2 Corinthians, Hebrews, Galatians, 
Philippians, Ephesians, 1 and 2 Thessalonians, and Colossians. 

Editions: [George Horner] The Coptic Version of the New Testament in 
the Northern Dialect, Otherwise Called Memphitic and Bohairic . . . , 4 vols. 
(Oxford, 1898-1905); idem, The Coptic Version of the New Testament in the 
Southern Dialect, Otherwise Called Sahidic and Thebaic . . . , 7 vols. (Oxford, 
1911-24) [each with a literal English translation]; Henri Hyvernat, Bybliothe- 
cae Pierpont Morgan Codices Coptici, photographice expressi . . . (Rome, 
1922), 56 vols, in 63 [the contents are indexed in Winifred Kammerer, A 
Coptic Bibliography (Ann Arbor, MI, 1950), pp. 33 f.]; Herbert Thompson, 
The Gospel of St. fohn According to the Earliest Coptic Manuscript (London, 
1924) [sub-Achmimic dialect]; idem, The Coptic Version of the Acts of the 
Apostles and the Pauline Epistles in the Sahidic Dialect (Cambridge, 1932) 
[Chester Beatty MSS.l; Rodolphe Kasser, Evangile de Jean et Genese J-1V, 2 
(Louvain, 1958) [Bodmer MS.]; Elinor M. Husselman, The Gospel of John in 
Fayumic Coptic (P. Mich. inv. 352 T) (Ann Arbor, MI, 1962); Das Mdttaus- 
Evangelium im Mitteldgyptischen Dialekt des Koptischen {Codex Scheide), 

100. For other noteworthy readings, see Metzger, "An Early Coptic 
Manuscript of the Gospel According to Matthew" in Studies in New Testa- 
ment Language and Text, ed. by J. K. Elliott (Leiden, 1976), pp. 307 f. 

101. For an account of preliminary textual analyses, see Metzger, 
Early Versions, pp. 118 f. An edition was published by Hans-Martin 
Schenke in vol. cxxxvii of Texte und Untersuchungen (Berlin, 199D- 

102. TitoOrlandi, Lettere di san Paolo in copto-ossirinchita (Milan, 1974). 

Important Witnesses to the Text of the New Testament 115 

ed. by Hans-Martin Schenke (Berlin, 1981); Liste der Koptischen Hand- 
schriften des Neuen Testaments; I. Die Sahidischen Handschriften der 
Evangelien, ed. by Franz-Jiirgen Schmitz and Gerd Mink (Berlin, 1986); 
Apostelgeschichte 1, 1-153 im mitteldgyptischen Dialekt des Koptischen 
(Codex Glazier), ed. by Hans-Martin Schenke (Berlin, 1991); H. Forster, 
"Papyrusfragmente eines sahidischen Corpus Paulinum," Zeitschrift fiXr 
Antikes Christentum, v (2001), pp. 3-22; Das Verhdltnis der Koptischen zur 
Griechischen Oberlieferung des Neuen Testaments; Dokumentation und 
Auswertung der Gesamtmaterialien beider Traditionen zum Jakobushrief 
und den beiden Petmsbriefen, ed. by Franz-Jurgen Schmitz (Berlin, 2003). 

4. The Gothic Version 103 

Shortly after the middle of the fourth century, Ulfilas, often 
called the "apostle to the Goths," translated the Bible from Greek 
into Gothic. For this purpose, he created the Gothic alphabet and re- 
duced the spoken language to written form. The Gothic version is 
the earliest known literary monument in a Germanic dialect. 

The most nearly complete of the half-dozen extant Gothic man- 
uscripts (all of which are fragmentary) is a deluxe copy dating from 
the fifth or sixth century and now in the Uppsala University Library. 
It contains portions of all four Gospels, which stand in the so-called 
Western order (Matthew, John, Luke, and Mark). It is written on pur- 
ple vellum in large letters of silver ink, whence the name that is com- 
monly given to this manuscript, Codex Argenteus, i.e., the "silver 
codex" (Fig. 20). The initial lines of the Gospels and the first line of 
every section of text are in gold letters. All the other manuscripts of 
the Gothic New Testament, with the exception of a vellum leaf from 
a bilingual Gothic-Latin codex, 104 are palimpsests. 

103. See Elfriede Stutz, "Das Neue Testament in gotischer Sprache," in 
Die alten Obersetzungen des Neuen Testaments, die Kirchenvdterzitate und 
Lectionare, ed. by Kurt Aland (Berlin, 1972), pp. 375—402; Metzger, Early 
Versions, pp. 375-88; J. N. Birdsall, "Gothic Versions," Anchor Bible 
Dictionary, vi (New York, 1992), pp. 803-5; B. M. Metzger, The Bible in 
Translation: Ancient and English Versions (Grand Rapids, MI, 2001), 

pp. 38-40; R. Gryson, "La version gothique des evangiles. Essai de 
reevaluation," Revue theologique de Louvain (1990). 

104. Unfortunately, the leaf was completely ruined in 1945 by unde- 
tected seepage of water from the Lahn River into a bank vault in Giessen, 
where it had been stored for safekeeping during the Second World War. 
Earlier in the twentieth century, P. Glaue and K. Helm had published a 
rather unsatisfactory photograph of the leaf (Zeitschrift fur die neutesta- 
mentliche Wissenschaft, xi [1910J, pp. 1-38). 


The Text of the New Testament 




test* 1 i;kltmfcMfcftTTM*> kkovI> 

n>.\m|n;niMN' is*.hnmnhi'H» 

i>n it Me jti^NiN^M km »M vxoihh 

A. AMXIA&JkMMMI&eMltl v ■».-•».» 
•' S'.KMMlMi:>N.V|:iAl\5MlMHK 

* i;|ihY^s^niM^HMceiti.mkhsM 

i n m \ ^ m * »i \Y a v r>H in i s ji^U 
&^in^:iaihii<}>mi.'iK' *in*JJi 
: >Mh-r*|CHOiu*)tbTnM!SrrhK » 


Figure 20 Gothic Codex Argenteus (fifth or sixth century), University 
Library, Uppsala; Mark 5.18-24. Actual size 9% X 7% inches. 

Ulfilas' translation is remarkably faithful to the original, fre- 
quently to the point of being literalistic. For the basis of his version, 
Ulfilas used that form of Greek text current in Byzantium about 
A.D. 350, belonging to the early Koine type of text. Western read- 
ings, particularly in the Pauline Epistles, were subsequently intro- 
duced from Old Latin manuscripts. 

Editions: Wilhelm Streitberg, Die gotiscbe Bibel, 2te Aufl. (Heidelberg, 
1919; repr. with additions, 5th ed. 1965) [Gothic and reconstructed Greek 
text on opposite pages]; O. van Friesen and A. Grape, eds., Codex Argenteus 
Upsaliensis, fussu senatus universitatis phototypice editus (Uppsala, 1927). 

Important Witnesses to the Text of the New Testament 117 

5. The Armenian Version 105 

Sometimes called "the queen of the versions," the Armenian ver- 
sion is generally regarded as one of the most beautiful and accurate of 
all early translations of the Bible (Fig. 21). With the exception of the 
Latin Vulgate, more manuscripts of this version are extant than of any 
other early version; Rhodes has catalogued 1,244 copies of all or part 
of the New Testament, 106 and it is known that several hundred more 
are in libraries within the former Soviet Union. Traditions differ regard- 
ing its origin. According to Bishop Koriun (died c. 450) and the histo- 
rian Lazar of Pharb (c. 500), it was St. Mesrop (died A.D. 439), a soldier 
who became a Christian missionary, who created a new alphabet and, 
with the help of the Catholicus Sahak (Isaac the Great, 390-439), trans- 
lated the version from the Greek text. On the other hand, Moses of 
Chorion, the nephew and disciple of St. Mesrop, says that Sahak made 
it from the Syriac text. Both views, with various modifications, have 
found defenders among modern scholars. There is some reason to 
think that the earliest Armenian version of the Gospels circulated in the 
form of a harmony, distantly related to Tatian's Diatessaron. 

The earliest Armenian version appears to have undergone a re- 
vision prior to the eighth century. Whether the Greek text that served 
as the basis for the revision was predominantly Caesarean or Koine 
in type is a question that has not yet been satisfactorily answered. In 
any case, the text of Matthew and Mark in many Armenian manu- 
scripts and even in Zohrab's printed edition appears to be strongly 
Caesarean in character. 

Editions: There is no satisfactory critical edition; the edition most fre- 
quently used was prepared by the Mechitarist Yovhan Zohrabian (Venice, 
1789; whole Bible 1805). The oldest known manuscript, a tetraevangelium 
copied A.D. 887, was photographically reproduced by G. Khalatheants 
(Moscow, 1899). See also Henning Lehmann and J. J. S. Weitenberg, eds., 
Armenian Texts, Tasks, and Tools (Aarhus, 1993); and Text and Context, 

105. For discussions of research on the Armenian version, see Louis 
Leloir, "La version armenienne du Nouveau Testament," in Die alten Uber- 
setzungen des Neuen Testaments, die Kirchenvdterzitate und Lektionare, 
ed. by K. Aland (Berlin, 1972), pp. 300-13; Joseph M. Alexanian, "The 
Armenian Version of the New Testament," in The Text of the New Testa- 
ment in Contemporary Research, ed. by Ehrman and Holmes, pp. 157-72. 

106. Erroll F. Rhodes, An Annotated List of Armenian New Testament 
Manuscripts (Tokyo, 1950). 


The Text of the New Testament 

|,MKJ1U1|^I.(M>I-Ilkl I; 
• rifln.'l M MriM*t4M>.f» 

.-< l»i'i.iM.iK./ri4Hir. y h 

I i:n;ui)i -mix pn . n, 

I'l.Ull.M.- l»m>UMIl;*f 

**sju*mBv\, -pi.. urW*. 

^"- f ^' i »■ — ^— -™ »^ ^W# ^T"* 

I' iTta-wm k»r 
VurtHunu iriri-'inoi: 

« t 



Figure 21 Armenian Gospel MS. 229 of the Patriarchal Library of 
Etchmiadzin (dated A.D. 989), now in the State Repository of Manuscripts, 
Erevan; mention of Ariston the Presbyter, col. b, between lines 6 and 7 
opposite decorative boss (see p. 325). Actual size of entire folio 13% X 
10% inches (plate reproduces lower three-quarters of folio 110 v ). 

Studies in the Armenian New Testament. Papers Presented to the Conference 
on the Armenian New Testament, May 22-28, 1992, ed. by S. Ajamian and 
M. E. Stone (Atlanta, 1994). 

6. The Georgian Version 

Of all the early versions of the New Testament, probably the 

: .mown among Western schc ars .; the Jt >rgian version. The 
t 1 'pie of Caucasian Georgia, that rough, mountainous district be- 
tween 'he Black Sea and the Caspian Sea, received the Gospel dur- 

„ie first half of the fourth century. The time and circumstances of 
the translation of the New Testament into Georgian, an agglutinative 
an'ia >e not known to be related to any other, are hidden in the 
mists of legend. Like the Armenian version, it is an important witness 
to the Caesarean type of text. 

Among the oldest known Gospel manuscripts are the Adysh 
manuscript of A.D. £97, thn Opiza manuscript of 913 and the Tbet' 

Important Witnesses to the Text of the New Testament 119 

manuscript of 995. In most apparatus critici, the Adysh manuscript is 
cited as Geo 1 and the testimony of the other two, as Geo 2 (A and B). 

Editions and Studies: The Old Georgian Version of the Gospel of Mark . . . , 
ed. with Latin trans, by Robert P. Blake (Paris, 1929); Matthew (Paris, 1933); 
John, ed. by Blake and Maurice Briere (Paris, 1950); F.uke, ed. by Briere (Paris, 
1955); Joseph Molitor, "Das Adysh-Tetraevangelium. Neu ubersetzt und mit 
altgeorgischen Paralleltexten verglichen," Oriens Christianus, xxxvii (1953), 
pp. 33-55; xxxviii (1954), pp. 11-40; xxxix (1955), pp. 1-32; xl (1956), pp. 
1-15; xli (1957), pp. 1-21; xlii (1958), pp. 1-18; xliii (1959), pp. 1-16; G. 
Garitte, I.'ancienne version georgienne des Actes des Apotres d'apres deux 
manuscritsde Sinai (Louvain, 1955). Other monographs and editions include 
Ilia Abuladze, The Acts of the Apostles According to the Old Manuscripts (in 
Georgian; Tiflis, 1950); K'et'evan Lort'k'ip'anidze, The Georgian Versions of 
the Catholic Epistles According to Manuscripts of the 10th to 14th Centuries 
(in Georgian; Tiflis, 1956), trans, into Latin by Joseph Molitor in Oriens 
Christianus, xlix (1965), pp. 1-17; 1 (1966), pp. 37—45; Ilia Imnaisvili, The 
Apocalypse of John and Its Commentary [i.e., the Commentary of Andrew of 
Caesareal (in Georgian; Tiflis, 1961), trans, into Latin by Joseph Molitor in 
Oriens Christianus, 1 (1966), pp. 1-12 (continued). For the Gospels, see J. 
Molitor, Synopsis latina Evangeliorum Ihericorum antiquissimorum secun- 
dum Matlhaeum, Marcutn, Lucam, desumpta e codicibus Adysh, Opiza, Tbeth 
necnon efragmentis hihlicis et patristicis quae dicuntur Chanmeti et Haemeti 
(Louvain, 1965). See also J. N. Birdsall, "Georgian Studies and the New Testa- 
ment," New Testament Studies, xxix (1983), pp. 306-20; idem, "The Georgian 
Versions of the Acts of the Apostles," in Text and Testimony, Essays on the New 
Testament in Honour of A. F.J. Klijn, ed. by T Baarda et al. (Kampen, 1988), 
pp. 39—45; idem, "The Georgian Version of the New Testament," in The Text 
of the New Testament in Contemporary Research: Essays on the Status Quaes- 
tionis, ed. by Bart D. Ehrman and Michael W. Holmes (Grand Rapids, MI, 
1995), pp. 173-87; L. Kadzaia, Die dltesten Georgische Vier-Evangelien- 
Handscrift, Teil I, Prolegomena. Aus dem Georgischen iiberscizt von H. 
Greeven und M. Job, with an English summary (Bochum, 1989), pp. 80-2; 
J. W. Childers, "The Old Georgian Acts of the Apostles: A Progress Report," 
New Testament Studies, xlii (1996), pp. 55-74. 

7. The Ethiopic Version 


The Ethiopic version has been preserved in a section of Africa 
that, for long periods of time, had been virtually isolated from the 
rest of Christendom. Although none of the extant manuscripts of the 

107. See Rochus Zuurmond, "The Ethiopic Version of the New Testa- 
ment," in The Text of the New Testament in Contemporary Research, ed. by 
Ehrman and Holmes, pp. 142-56. 

120 The Text of the New Testament 

version is older than perhaps the tenth century and most of them 
date from the fifteenth and later centuries, 108 the version is of con- 
siderable interest on several scores. Until the latter part of the twen- 
tieth century, the status of the last 12 verses of Mark in this version 
was in doubt, arising from conflicting statements made concerning 
the evidence of the same manuscripts. Now, however, on the basis 
of the personal examination of photographs of the ending of the sec- 
ond Gospel in 65 Ethiopic manuscripts, belonging to about 30 dif- 
ferent collections, Metzger has ascertained that all of them have the 
text of Mark 16.9-20. In addition, what is known as the "shorter end- 
ing" of Mark, found in several Greek and Syriac manuscripts, occurs 
in many Ethiopic manuscripts between 16.8 and 9. Subsequently, 
William F. Macomber of the Hill Monastic Manuscript Microfilm 
Library at Collegeville, Minnesota, examined microfilms of 129 addi- 
tional Ethiopic manuscripts of Mark. Of the total of 194 (65 + 1 29) 
manuscripts, all but two (which are lectionaries) have Mark 16.9-20, 
while 131 contain both the shorter ending and the longer ending. 109 

At the end of the Ethiopic version of Acts chapter 28, there is a 
directive to readers to consult Paul's letters and the Acts of Paul, as 
well as information about Paul's further activities in Rome. This ad- 
dition originated as a gloss or colophon that was later integrated into 
the main text. 110 

Scholars differ on the question of the date of origin of the 
Ethiopic version; some argue for a date as early as the fourth century, 
while others attribute it to the sixth or seventh century. Opinion also 
differs as to whether the translators made use of a Greek or Syriac 
original. In any case, it is a curious fact that in the Epistles of Paul the 
version frequently agrees with p 46 , with little or no other support. 

108. For paleographical studies, see Siegbert Uhlig, "Probleme, der 
athiopischen Palaographie," Zeitschrift der deutschen morgenldndiscben 
Gesellschaft, Supplement v (1983), pp. 135—45; D. M. Davies, "The Dating 
of Ethiopic Manuscripts," Journal of Near Eastern Studies, xlvi (1987), 
pp. 287-307. 

109. For the Ethiopic text and English translation of the shorter 
ending, see Metzger's New Testament Studies, Philological, Versional, and 
Patristic (Leiden, 1980), pp. 127-47; for a critical edition of Mark, see 
Rochus Zuurmond, Novum Testamentum Aethiopice; The Synoptic Gospels, 
part I (Stuttgart, 1989). 

110. Siegbert Uhlig, "Ein pseudepigraphischer Actaschluss in der 
athiopischen Version," Oriens Christianus, lxxiii (1989), pp. 129-36. 

Important Witnesses to the Text of the New Testament 121 

The version also shows evidence of later contamination from Coptic 
and Arabic texts. Thus, the Ethiopic text eventually became a con- 
glomerate with quite disparate elements standing side by side. 
Analyses of the earlier form of the Ethiopic version disclose a mixed 
type of text, predominantly Byzantine in complexion but with occa- 
sional agreement with certain early Greek witnesses (p 46 and B) 
against all other witnesses. The little that is known of this version as 
far as the New Testament is concerned (the Old Testament has been 
studied more thoroughly) suggests that it deserves far more attention 
than it has received. The earliest known manuscript, a codex of the 
four Gospels, dates from the thirteenth century; most other manu- 
scripts are of the fifteenth and succeeding centuries. 

Editions: The editio princeps of the Ethiopic New Testament was pub- 
lished by three Abyssinian monks, who issued their work in two volumes 
under the pseudonym Petrus Ethyops (Rome, 1548-9). This text was 
reprinted, with a Latin translation, in Brian Walton's Polyglot Bible (1657). 
Other more recent editions, made for modern missionary purposes, were 
edited by T. Pell Piatt (London, 1826), re-edited by F. Pratorius (Leipzig, 
1899, repr. 1914), and F. da Bassano (Asmara, 1920). The readings of the 
editio princeps are referred to by using the siglum Eth ro . 

8. The Old Slavonic Version 111 

With the exception of St. Jerome, more is known of the life and 
work of SS. Cyril and Methodius, the apostles to the Slavs, than of any 
other translators of an ancient version of the Bible. Sons of a wealthy 
official in Salonica, they are credited with the creation of the 
Glagolitic alphabet, as well as the so-called Cyrillic alphabet. Soon 
after the middle of the ninth century, they began translating the 
Gospels (probably in the form of a Greek lectionary) into Old Bul- 
garian, commonly called Old Slavonic. The version belongs basically, 
as one would expect, to the Byzantine type of text, but it also con- 
tains not a few earlier readings of the Western and Caesarean types. 

111. For surveys of research on the Old Slavonic version, see Metzger, 
Chapters in the History of New Testament Textual Criticism (Grand Rapids, 
MI, 1963), pp. 73-96; Christian Hannick, "Das Neue Testament in 
altkirchenslavischer Sprache," Diealten Ubersetzungen desNeuen Testa- 
ments, die Kirchenvaterzitate und Lektionare, ed. by K. Aland (Berlin, 1972), 
pp. 403-35; and Marcello Garzaniti, Die altslauische Version der Evangelien: 
Forschungsgeschichte und zeitgenossische Forschung (Cologne, 2001). 

122 The Text of the New Testament 

Soon after Cyril and Methodius prepared their version, it was 
subjected to the process of emendation and adaptation in accor- 
dance with the dialectal needs of the various Slavic peoples by 
whom it was used. All extant manuscripts, which date from the 
eleventh century onward, embody, to a greater or lesser extent, re- 
censions that are characterized by Bulgarian, Czech, Russian, or Ser- 
bian peculiarities. The Slavonic Bible also served as the basis for 
early translations into Czech (1475) and Bulgarian (1840) and influ- 
enced versions in the other Slavic languages. 

Preliminary work of classifying Slavonic manuscripts was begun 
at the end of the nineteenth century by Professor G. N. Voskresenski 112 
of Moscow, who divided them into four groups that represent, as he 
thought, four families. The oldest recension is preserved in the South 
Slavic manuscripts, to which most of the famous codices belong. The 
second recension is preserved in the oldest Russian manuscripts, dat- 
ing from the eleventh and twelfth centuries. The other two recensions 
belong to the fourteenth and fifteenth centuries. The Apocalypse 
stands in a class by itself, not having been translated, according to a 
monograph by Oblak, 113 until the twelfth century. 

Editions: Josef Vajs, Evangelium sv. Matouse, text rekonstruovany 
(Prague, 1935) [with the reconstructed underlying Greek text printed on op- 
posite pages]; . . . Marka (Prague, 1935); . . . Lukdse (Prague, 1936); . . .Jana 
(Prague, 1936). 

9. Other Ancient Versions 

a. The Arabic Versions 

The earliest translations of the Gospels into Arabic probably date 
from the eighth century. According to a survey made by Ignazio 
Guidi, 114 the manuscripts of these Arabic translations fall into five 

112. Data on the publications of Voskresenski may be found in Metzger, 
Chapters in the History, p. 78, n. 5. 

113. V. Oblak, "Die kircheslavische Ubersetzung der Apokalypse," 
Archiv fur slavische Philologie, xiii (1891), pp. 321-61. 

114. "Le traduzioni degli Evangelii in arabo e etiopico," Atti clella R. 
Accidentia dei Lincei, Memorie, anno cclxxi, serie quarta, classe di scienze 
morali, storiche e filolgiche, iv, part I (Rome, 1888), pp. 5-76. See also 
Samis Arbache, "Les versions arabes des Evangiles," Melanges de science 
religieuse, lvi, no. 3 (1999), pp. 85-93. 

Important Witnesses to the Text of the New Testament 123 

main groups: (1) those made directly from the Greek; (2) those made 
directly from or corrected from the Syriac Pershitta; (3) those made 
directly from the Coptic, whether Sahidic, Bohairic, or some other 
dialect; (4) those of two distinct eclectic recensions produced by the 
Alexandrian patriarchate during the thirteenth century; and (5) mis- 
cellaneous manuscripts, some of which are characterized by being 
cast into the form (if rhymed prose, made classic by the Qur'an. Also, 
more than one Arabic version has been corrected by others derived 
from a different basic text. 

From the Middle Ages to the nineteenth century, still other Arabic 
translations of parts of the Bible were made for various ecclesiastical 
groups as well as in a variety of forms of Arabic. The former include 
renderings made for Melchites, Maronites, Nestorians, Jacobites, and 
Copts; the latter include, besides classical Arabic, those forms of the 
language currently used in Algeria, Chad, Egypt, Morocco, Palestine, 
Sudan, and Tunisia, as well as the vernacular of Malta. 

b. the Old Nubian Version 

During the early centuries of the Christian era, Nubia, which lies 
between Egypt on the north and Ethiopia on the south, comprised 
three independent kingdoms. When Christianity first reached the 
Nubian people is not known, but the vast stretches south of Egypt 
would have given shelter to more than one Christian driven from Egypt 
by the persecutions ordered by the Roman Emperor Diocletian in 303- 
The first formally designated missionaries arrived in Nubia about 
the middle of the sixth century. 115 Over the centuries, the number of 
congregations in Nubia multiplied and were counted, we are told, by 
the hundreds. For about five centuries, Christianity flourished, provid- 
ing the chief cohesive element in Nubian society. By the close of the 
fourteenth century, however, having been cut off from the rest of the 
Christian world by Arab invaders pressing southward from Muslim 
Egypt, the weakened Nubian Church was ready to expire. The grow- 
ing power of Arabs hemmed in the Nubian Christians on the north, 
east, and west and, finally, the whole population embraced Islam. 

115. See B. M. Metzger, "The Christianization of Nubia and the Old 
Nubian Version of the New Testament," in his volume Historical and 
Literary Studies, Pagan, Jewish, and Christian (Grand Rapids, MI, 1968), 
pp. 112-22; Piotr O. Scholz, "Nubisches Christentum im Lichte seinen 
Wandmalereinen," Oriens Christianus, lxxxiv (2000), pp. 178-231. 

1 24 The Text of the New Testament 

It was only in the twentieth century that evidence for the Old 
Nubian version came to light. In 1906, Dr. Karl Schmidt purchased in 
Cairo a quire of 16 multilated pages from a parchment codex ac- 
quired in Upper Egypt. This contained a portion of a lectionary for 
Christmastide, corresponding to 20-26 December. 116 For each day, a 
section of Scripture is supplied from the Apostolos (Romans, 
Galatians, Philippians, and Hebrews) and from the Gospel (Matthew 
and John). Like other texts in Nubian, the lectionary is written in an 
alphabet that is essentially Coptic, reinforced by additional letters 
needed to represent sounds peculiar to the language. Toward the 
end of the twentieth century, other biblical fragments in Old Nubian 
came to the attention of scholars. These include passages from the 
Gospel of John and the Book of Revelation. 117 

c. The Sogdian Version 

The Sogdian language, a Middle Iranian tongue, was an eastern 
member of the Indo-European family of languages. During the sec- 
ond half of the first millennium of the Christian era, Sogdian was 
widely used in East Turkestan and adjacent areas of Central Asia. Be- 
cause of the Sogdians' energetic pursuit of colonization and trading 
activity, documents in their language were carried far and wide. 

In contrast to what is now known to have been the widespread 
dissemination of Sogdian, before the twentieth century scholars 
knew very little about the language. Then, in 1903, a wide variety of 
Sogdian manuscripts came to light near Turfan; and when the lan- 
guage had been deciphered, it was found that, in addition to exten- 
sive remains of Manichean and Buddhist texts, there were a number 
of Christian documents written in a purely consonental script 
resembling Estrangela Syriac. These proved to be accounts of sev- 
eral martyrdoms; a portion of the Shepherd of Hermas; fragments 

116. Most recently re-edited by Gerald M. Browne, Griffith's Old 
Nubian Lectionary (Rome, 1982). 

117. For the text of all the known fragments of the Old Nubian ver- 
sion, see Gerald M. Browne, Bibliorum sacrorum versio palaeonubiana 
(Louvain, 1994). The same scholar has also produced the Old Nubian Dic- 
tionary (Louvain, 1996), in which he provides (whenever they are avail- 
able) the Greek Voriage and any parallel Coptic version, as well as adding 
cognates from the modern Nubian dialects. 

Important Witnesses to the Text of the New Testament 125 

preserving a considerable number of passages of Matthew, Luke, 
and John; and a few small scraps of 1 Corinthians and Galatians. 118 
It is thought that translations of Christian documents began to be 
made following the vigorous Nestorian mission in Central Asia dur- 
ing the seventh century. lly 

d. The Anglo— Saxon Versions 

The oldest extant New Testament material in Anglo-Saxon is in 
the form of interlinear glosses on a Vulgate Latin text, for glossing 
was a form of early English pedagogy. While such glosses give a 
rendering of (almost) every word in the text, they naturally follow 
the Latin order instead of the Anglo-Saxon and would therefore re- 
quire rearrangement before they could serve as an independent 
translation. In fact, they were not intended to take the place of the 
Latin text but to help the reader understand it. 

The famous Lindisfarne Codex (see p. 108) was written about the 
year 700 by Eadfrid, bishop of Lindisfarne. During the tenth century, 
the manuscript was transferred to Chester-le-street, near Durham; 
and there, before 970, an interlinear gloss in the rather rare literary 
dialect of Old Northumbrian was added under the supervision of 
a priest named Aldred. Another famous Latin manuscript, the 
Rushworth Gospels, written in an Irish hand of about 800, was 
glossed in the latter half of the tenth century. The gloss in Matthew is 
the rare Old Mercian dialect, which was spoken in the central part of 
England, while that added to the other three Gospels is a form of Old 

Six other manuscripts contain glossing in a West Saxon form of 
Anglo-Saxon used in Wessex. In some cases, the West Saxon read- 
ings differ from the Vulgate but agree with the bilingual Codex 
Bezae (see pp. 70-3), while other readings represent a misunder- 
standing of the Latin. From what has been said thus far, the limita- 
tions of the Anglo-Saxon version in representing the Greek text of 

118. For popular accounts of the Sogdian materials, see J. Rendel 
Harris, Side-Lights on the New Testament Research (London, 1908), 

pp. 115-24; Louis H. Gray, 'New Testament Fragments from Turkestan," 
Expository Times, xxv (1913-14), pp. 59-61. 

119. See Ian Gilman and Hans-Joachim Klimkeit, Christians in Asia 
Before 1500 '(Ann Arbor, MI, 1999), pp. 207-21. 

126 Thk Text of the New Testament 

the Gospels will be obvious enough. The chief value of the version 
for the textual critic is its contributions to tracing the history of the 
Latin Bible. 120 

III. Patristic Quotations from 
the New Testament 

Besides textual evidence derived from New Testament Greek 
manuscripts and from early versions, the textual critic has available 
the numerous scriptural quotations included in the commentaries, 
sermons, and other treatises written by early Church fathers. 121 
Indeed, so extensive are these citations that if all other sources for 
our knowledge of the text of the New Testament were destroyed, 
they would be sufficient alone for the reconstruction of practically 
the entire New Testament. 

The importance of patristic quotations lies in the circumstance 
that they serve to localize and date readings and types of text in Greek 
manuscripts and versions. For example, since the quotations that 
Cyprian, bishop of Carthage in North Africa about A.D. 250, includes 
in his letters agree almost always with the form of text preserved in 
the Old Latin manuscript k, scholars have correctly concluded that this 

120. N. R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon 
(Oxford, 1957); Minnie Kate Morrell, A Manual of Old English Biblical 
Materials (Knoxville, TN, 1965); M. Grilnberg, The West-Saxon Gospels: 
A Study of the Gospel of St. Matthew, with the Text of the Four Gospels 
(Amsterdam, 1967); R. M. Liuzza, ed., The Old English Version of the 
Gospels, vol., 1, Text and Introduction (Oxford, 1994), vol. 2, Notes and 
Glossary (Oxford, 2000); Richard .Marsden, '"Ask What I Am Called': The 
Anglo-Saxons and Their Bibles," in The Bible as Book: The Manuscript 
Tradition, ed. by John L. Sharp III and Kimberly Van Kampen (London. 
1998), pp. 145-76. 

121. For recent discussions of the significance of the patristic wit- 
nesses for the New Testament text, see the following contributions to The 
Text of the New Testament in Contemporary Research, ed. by Ehrman and 
Holmes: Gordon D. Fee, "The Use of the Greek Fathers for New Testa- 
ment Textual Criticism," pp. 191-207; J. Lionel North, "The Use of the 
Latin Fathers for New Testament Textual Criticism," pp. 208-23; Sebastian 
P. Brock, "The Use of the Syriac Fathers for New Testament Textual 
Criticism," pp. 224-36. 

Important Witnesses to the Text of the New Testament 127 

fourth- or fifth-century manuscript is a descendant of a copy current 
about 250 in North Africa. Occasionally, it happens that a patristic 
writer specifically cites one or more variant readings present in man- 
uscripts existing in his day. Such information is of the utmost impor- 
tance in providing proof of the currency of such variant readings at a 
given time and place. 122 

Before patristic evidence can be used with confidence, how- 
ever, one must determine whether the true text of the ecclesiastical 
writer has been transmitted. As in the case of New Testament man- 
uscripts, so also the treatises of the fathers have been modified in 
the course of copying. The scribe was always tempted to assimilate 
scriptural quotations in the fathers to the form of text that was cur- 
rent in the later manuscripts of the New Testament, a text that the 
scribes might well know by heart. 123 When the manuscripts of a 
father differ in a given passage, it is usually safest to adopt the one 
that diverges from the later ecclesiastical text (the Byzantine Text or 
the Vulgate). 

After the text of the patristic author has been recovered, the fur- 
ther question must be raised of whether the writer intended to quote 
the scriptural passage verbatim or merely to paraphrase it. If one is 
assured that the father makes a bona fide quotation and not a mere 

122. For discussions of such explicit references in the works of Origen 
and Jerome, reference may be made to Bruce M. Metzger, Historical and 
Literary Studies, pp. 85-103 (on Origen), and idem, New Testament Studies, 
pp. 199-210 (on Jerome). 

123. The requirements of memorizing portions of the Scriptures for 
ordination to the deaconate and the priesthood are specified in a Coptic 
ostracon edited by Crum. According to the ostracon, Samuel, Jacob, and 
Aaron, who applied to Bishop Abraham to be ordained as deacons, were 
required "to master the Gospel according to John and learn it by heart by 
the end of Pentecost and to recite it." Aphou, bishop of Oxyrhynchus, is 
said to have required a deacon at ordination to know 25 Psalms, two Kpis- 
tles of Paul, and a portion of a Gospel by heart; a priest had to know, in 
addition, portions of Deuteronomy, Proverbs, and Isaiah (W. K. Crum, 
Coptic Ostracafrom the Collections of the Egypt Exploration Eund [London, 
1902], p. 9, n. 29). According to the rules of St. Pachomius, applicants for 
entrance into the monastery were required to know 20 Psalms, two Kpis- 
tles of Paul, or a portion of some other part of Scripture (Regulae Monasti- 
cae S. Pachomii, ed. by P. B. Albers, p. 41; see also Richard Reitzenstein, 
Historia Monachorum, pp. 6l f., 162 ff.). 

128 The Text of the New Testament 

allusion, the problem remains of whether he quoted it after consult- 
ing the passage in a manuscript or relied on his memory. 124 The for- 
mer is more probable in the case of longer quotations, whereas 
shorter quotations were often made from memory. Furthermore, if the 
father quotes the same passage more than once, it often happens that 
he does so in divergent forms. 

Despite the difficulties that attend the determination and evalua- 
tion of patristic sources, this kind of evidence can be of great impor- 
tance in establishing the oldest form of the text. In Luke 3.22, for 
example, the great majority of witnesses, starting with p 4 , indicate 
that at Jesus' baptism the voice from heaven proclaimed "You are my 
beloved Son, in you I am well pleased" (in reference to Isa. 42.1). 
Other witnesses, however, indicate that the voice instead quoted 
Psalm 2.7: "You are my Son, today I have begotten you." The earli- 
est Greek manuscript to record the words in this latter form is the 
fifth-century Codex Bezae, but the text is also quoted numerous 
times in patristic sources of the second to fourth centuries, including 
Justin, Clement of Alexandria, Origen, Methodius, the Gospel ac- 
cording to the Hebrews, the Gospel according to the F.bionites, and 
the Didascalia. In all these cases, it is the alternative form, that found 
in Codex Bezae, that is quoted. This means that, with the exception 
of p 4 , all of the surviving witnesses of the second and third centuries 
appear to have known this alternative form of the text. 

Patristic evidence is valuable not only for establishing the earli- 
est form of the text but even more for tracing the history of the text's 
transmission. For example, full analyses of such Alexandrian wit- 
nesses as Clement, Origen, Didymus, Athanasius, and Cyril can show 
how the text of the New Testament developed in Lower Egypt over 
the first 400 years. 

A special kind of information supplied by Church fathers is 
the explicit references they occasionally make concerning variant 
readings among the manuscripts known to them in their own day. 
Sometimes the father will also express his opinion as to the value of 

124. Even when an author makes an explicit reference to the text he 
is quoting, there is no guarantee that he had a text to hand. In a passage 
of Didymus the Blind's commentary on the Psalms, e.g., he indicates that 
Matthew and Luke had slightly different wordings for a passage they had 
in common; unfortunately, Didymus confused the two references, attribut- 
ing the text of Luke 11.13 to Matthew and the text of Matthew 5.11 to 
Luke! See Bart D. Khrman, Didymus the Blind and the Text of the Gospels 
(Atlanta, 1986) p. 9, n. 16. 

Important Witnesses to the Text of the New Testament 129 

this or that reading, 125 an opinion that may not be shared by modern 
scholars. In any case, it is of considerable significance when the 
father mentions that in his day such and such a reading was current 
in "few" or "many" or "most" manuscripts. This information may — or, 
more likely, may not — be matched by the survival today of copies 
containing the reading in question. 126 Such a comparison of ancient 
testimony with what is now extant discloses how very often chance 
has entered into the survival or the loss of manuscript evidence that 
existed in previous centuries. 

Given the significance of the patristic sources for establishing the 
oldest form of the text and for writing the history of its transmission, 
it is obviously important to have complete and reliable presentations 
of all the quotations of and allusions to the New Testament text in 
the church writers from the early centuries of Christianity. A new 
monograph series has been established to meet that need. Originally 
edited by Gordon Fee, The New Testament in the Greek Fathers: Texts 
and Analyses {now published by Society of Biblical Literature Press) 
is intended to devote separate volumes to individual fathers, either 
for their complete text of the New Testament or for a portion of it. 
To date, volumes have appeared on Didymus the Blind (the 
Gospels), Gregory of Nyssa (entire New Testament), Origen (sepa- 
rate volumes on the fourth Gospel and 1 Corinthians), and Cyril of 
Jerusalem (the entire New Testament). Volumes expected to appear 
soon will cover Athanasius (the Gospels), Basil the Great (Matthew), 
and Epiphanius (Acts, Epistles, and Revelation). 127 

125. See the chapter "The Practice of Textual Criticism Among the 
Church Fathers" in Metzger's New Testament Studies, pp. 189-98. 

126. For discussions of such explicit references in the works of Origen 
and Jerome, reference may be made to Metzger's Historical and Literary 
Studies, pp. 85-103 (on Origen), and New Testament Studies, pp. 199-210 
(on Jerome). Although James E. G. Zetzel collects some interesting data in his 
dissertation "Latin Textual Criticism in Antiquity" (New York, 1981), he omits 
(as do many other classical scholars) any consideration of the Church fathers. 

127. Ehrman, Didymus the Blind; James A. Brooks, The New Testa- 
ment Text of Gregory of Nyssa (Atlanta, 1991); Bart D. Ehrman, Gordon D. 
Fee, and Michael W. Holmes, The Text of the Fourth Gospel in the Writings 
of Origen (Atlanta, 1992); Darrell D. Hannah, The Text of 1 Corinthians in 
the Writings of Origen (Atlanta, 1997); Roderic L. Mullen, Ihe New Testa- 
ment Text of Cyril of Jerusalem (Atlanta, 1997). The forthcoming volumes in 
the series are by John Brogan on Athanasius, Jean-Francois Racine on Basil 
the Great, and Carroll Osburn on Epiphanius. 

130 The Text of ti if. New Testament 

Each volume in the series contains (1) a brief introduction, indi- 
cating the father's historical and theological context; (2) a complete 
listing of all New Testament citations, adaptations, and allusions 
found in his writings; (3) full collations of these materials against 
leading manuscript witnesses of the different textual families; 128 
(4) detailed analyses of the agreements and disagreements of the 
father's quotations with these various textual witnesses, usually fol- 
lowing the quantitative method for establishing textual relations and 
some kind of group profile form of analysis; 129 and (5) conclusions as 
to the significance of these data for understanding the history of the 
transmission of the text of the New Testament at the time and place in 
which the father lived. 

Significant conclusions have been reached by several of these 
studies. For example, James Brooks' study of Gregory of Nyssa, Jean- 
Francois Racine's study of Basil the Great, and Rod Mullen's study of 
Cyril of Jerusalem have shown that each of these fathers represents 
an early form of the Byzantine textual tradition, pushing the earliest 
stages of the Byzantine text type back into the late fourth century. At 
the same time, Bart Ehrman's study of the quotations of Didymus the 
Blind showed that this important fourth-century father was a solid 
witness to the Alexandrian type of text; even more, it showed that 
we must reevaluate the nature of the Alexandrian tradition. No 
longer is it appropriate to think in terms of a distinct form of "late 
Alexandrian" witness. Instead, all of the witnesses related to the 
Alexandrian text appear to attest it in greater or lesser purity, so in- 
stead of "early" and "late" Alexandrian texts, we have "primary" and 
"secondary" witnesses, depending on the relative proximity of each 
witness to the distinctively Alexandrian form of text. 

Clearly, the patristic sources are among the least studied but 
most important witnesses to be analyzed. The following is a list of 
several of the more important Church fathers whose writings contain 
numerous quotations from the New Testament. 130 

128. See pp. 306 ff. 

129. See pp. 236 ff. 

130. For information concerning the literary contributions of these 
fathers as well as bibliographical references to editions and monographs, 
see Johannes Quasten, Patrology (4 vols. Westminster, MD, 1950 ff.) and 
more recently Siegmar Dopp and Wilhelm Geerlings, eds., Dictionary of 
Early Christian Literature, trans, by Matthew O'Connell (New York, 2000). 

Important Witnesses to the Text of the New Testament 131 

Ambrose of Milan, d. 397 

Ambrosiaster (= pseudo- Ambrose) of Rome, second half of fourth 

Athanasius, bishop of Alexandria, d. 373 
Augustine, bishop of Hippo, d. 430 
Chrysostom, bishop of Constantinople, d. 407 
Clement of Alexandria, d. c. 212 
Cyprian, bishop of Carthage, d. 258 
Cyril of Alexandria, d. 444 
Didymus of Alexandria, d. c. 398 
Ephraem the Syrian, d. 373 
Epiphanius, bishop of Salamis, d. 403 
Eusebius, bishop of Caesarea, d. 339 or 340 
Gregory of Nazianzus in Cappadocia, d. 389 or 390 
Gregory of Nyssa in Cappadocia, d. 394 
Hilary of Poitiers, d. 367 
Hippolytus of Rome, d. 235 
Irenaeus, bishop of Lyons, d. c. 202 
Isidore of Pelusium, d. c. 435 
Jerome (= Hieronymus), d. 419 or 420 
Justin Martyr, d. c. 165 
Lucifer of Calaris (Cagliari), d. 370 or 371 
Marcion, flourished at Rome, c. 150-60 
Origen of Alexandria and Caesarea, d. 253 or 254 
Pelagius, fourth-fifth century 

Primasius, bishop of Hadrumentum, d. soon after 552 
Pseudo-Hieronymus, fifth-sixth century 
Rufinus of Aquileia, d. 410 
Tatian, flourished c. 170 
Tertullian of Carthage, d. after 220 
Theodore of Mopsuestia in Cilicia, d. 428 

Among these writers, one of the more controversial figures was 
Tatian, a Syrian from Mesopotamia, known chiefly for his Diatessaron, 
or Harmony of the four Gospels. Combining distinctive phrases pre- 
served by only one Evangelist with those preserved by another, he 
arranged the several sections of the Gospels into a single narrative. 
Omitting only a very few sections (such as the genealogies of Jesus in 
Matthew and in Luke, the former of which traces Jesus' lineage from 
Abraham onward and the latter of which traces it backward to Adam), 
Tatian managed to preserve practically the entire contents of the 

132 The Text of the New Testament 

separate Gospels woven into one. The name by which his work 
came to be known is Diatessaron, iil derived from the Greek phrase 
did reoadgcov, meaning "through [the] four [Gospels]." 

Tatian's harmony soon became popular, particularly in the East. 
As late as the fifth century, Theodoret, who became bishop of Cyrrhus 
or Cyrus on the Euphrates in upper Syria in A.D. 423, found that many 
copies of the Diatessaron were in use within his diocese. But because 
Tatian had become heretical in his later life and Theodoret believed 
that orthodox Christians were in danger of being corrupted by 
Tatian's work, he destroyed all of the copies of the Diatessaron that 
he could find (totaling about 200) and put in their place the separate 
Gospels of the four Evangelists. 132 

As a result of the zeal of Bishop Theodoret, and doubtless of 
others like him, no complete copy of Tatian's Diatessaron is extant 
today. In 1933, a small fragment of parchment containing a portion 
of a harmony in Greek was unearthed by archaeologists on the site 
of the ancient Roman fortress town of Dura-Europos on the lower 
Euphrates. It is known that this town fell to the Persians under 
King Shapur I in A.D. 256-7; therefore, the fragment must antedate 
that event. 133 The text that is preserved contains the narrative of the 
coming of Joseph of Arimathea for the body of Jesus. A literal trans- 
lation will show how words and phrases from all four Gospels have 
been woven together. Since the left-hand margin of the vellum has 
suffered damage, the first half-dozen or so letters at the beginning 
of each line are lacking. They can be restored, however, with al- 
most perfect confidence. In the following rendering, the restora- 
tions are enclosed within square brackets and the modern scriptural 

131- Victor of Capua's reference (in Codex Fuldensis; see p. 108) to 
Tatian's work as a diapente is not a lapsus calami, as some have thought, 
but is to be explained as the application to the harmony of a musical term 
that involves four intervals; see Franco Bolgiani, Vittore di Capua e il 
"Diatessaron" (Memorie dell' Accademia delle Scienze di Torino, Classe di 
scienze morali, storiche e filologiche, series 4 a , n. 2, 1962). 

132. Theodoret, Treatise on Heresies, i, 20. 

133. The fragment was edited by Carl H. Kraeling in Studies and 
Documents, vol. iii (London, 1935). The text was re-edited, with minimal 
emendations, by C. B. Wells in C. B. Wells, R. O. Funk, and J. F. Gilliam, 
The Parchments and Papyri: The Excavations at Dura-Europas . . . Final 
Report, v. 1, ed. by A. Perkins (New Haven, 1959), pp. 73^J. In the 
Milnster catalogue, the fragment has been assigned the number 0212. 

Important Witnesses to the Text of the New Testament 133 

references (which are not, of course, in the fragment) are enclosed 
within parentheses. 

[. . . the mother of the sons of Zebecllee (Matt. 27.56) and Salome 
(Mark 15.40) and the wives [of those who) had followed him from 
[Galilele to see the crucified (Luke 23.49b-c). And [the dajy was Prepa- 
ration; the Sabbath was daw[ning] (Luke 23.54). And when it was 
evening (Matt. 27.57), on the Preparation], that is, the day before the 
Sabbath (Mark 15.42), [there camel up a man (Matt. 27.57), be[ing] a 
member of the council (Luke 23.50), from Arimathea (Matt. 27.57), a 
c[i]ty of Qude]a (Luke 23-51), by name Jotseph] (Matt. 27.57), good and 
ri[ghteous] (Luke 23.50), being a disciple of Jesus, but se[cret]ly, for fear 
of the [Jewls (John 19.38). And he (Matt. 27.57) was looking for [the] 
kfingdom] of God (Luke 23.51b). This man [had] not [consented to 
[their] pturpose] (Luke 23.51a) . . . 

It is evident that Tatian went about composing his Diatessaron 
with great diligence. Probably, he worked from four separate manu- 
scripts, one of each of the Gospels; and as he wove together phrases, 
now from this Gospel and now that, he would no doubt cross out those 
phrases in the manuscripts from which he was copying. Otherwise, it 
is difficult to understand how he was able to put together so success- 
fully a cento of very short phrases from four separate documents. 

The most spectacular reading preserved in this fragment is near 
the beginning. Although it rests partly on a restoration and none of 
the secondary translations of Tatian that were known hitherto ex- 
hibits the reading, it is probable that Tatian referred to "the wives of 
those who had followed" Jesus from Galilee. This statement and the 
implications that it conveys are without parallel in the text of the sep- 
arate Gospels in any other manuscript or version. 

In addition to the Greek fragment, portions of Tatian's harmony 
are known through the quotations from it that certain early Syrian 
Church fathers included in their homilies and other treatises, particu- 
larly in the commentary that St. Ephraem of the fourth century wrote 
on the Diatessaron. Until the mid-twentieth century, this commentary 
was available only in an Armenian translation, preserved in two man- 
uscripts that were copied A.D. 1195. In 1957, announcement was 
made of the acquisition by Sir Chester Beatty of a Syriac manuscript 
containing about three-fifths of Ephraem's treatise. 134 The manuscript, 

134. For a description and a preliminary discussion of the manu- 
script, see Dom Louis Leloir, "L'Originale syriaque du commentaire de S. 
Ephrem sur le Diatessaron," Biblica, xl (1959), pp. 959-70 {Studia Biblica 

134 Thk Text of the New Testament 

which dates from the late fifth or early sixth century, was edited by 
Dom Louis Leloir, 135 who previously prepared a careful edition and 
translation of the commentary in its Armenian form. Several har- 
monies in other languages (Arabic and Persian 136 in the East and 
Latin, medieval Dutch, Old English, and Old Italian in the West) show 
more or less dependence, either in form or in text, upon Tatian's 
pioneering work. 

How much contamination the Diatessaron has exerted on the 
text of the separate Gospels has been variously estimated. Not many 
scholars today agree with the extreme views of Hermann von Soden 
and Anton Baumstark, who thought that they detected the presence 
of Tatianic influence on a great number of Eastern and Western man- 
uscripts of the Gospels. It is doubtless true, however, that not a few 
instances of harmonization of the text of the Gospels in certain wit- 
nesses (notably the Western witnesses) are to be ascribed to Tatian's 
influence. 137 

et Orientalia, ii, pp. 391-402). Tjitze Baarda has edited a brief quotation of 
Ephraem's commentary that is included in a Nestorian manuscript brought 
to Rome in 1869 ("A Syriac J ; ragment of Mar Ephraem's Commentary on 
the Diatessaron," New Testament Studies, viii [1962], pp. 287-300). 

135. Saint Ephrem, Commentaire de VEvangile concordant, texte syri- 
aque (Manuscrit Chester Beatty 709) (Dublin, 1963). See also L. Leloir, 
"Divergences entre l'original syriaque et la version armenienne du com- 
mentaire d'Ephrem sur la Diatessaron," in Melanges, Eugene Tisserant, ii 
(Vatican City, 1964), pp. 303-31. 

136. For a discussion of the art in the Persian Diatessaron, see Carl 
Nordenfalk, Art Bulletin, 1 (1968), pp. 119-40; Meyer Schapiro and Seminar. 
Art Bulletin, lv (1973), pp. 494-531; and Nordenfalk, op. cit., pp. 532-46. 

137. In addition to William L. Petersen's indispensable volume Tatian's 
Diatessaron: Its Creation, Dissemination, Significance, and History in 
Scholarship (New York, 1994), see T. Baarda, "Matthew 18.14c. An 'Extra 
Canonical' Addition in the Arabic Diatessaron?" Museon, cvii (1994), 

pp. 135-49; P. Joose, "An Introduction to the Arabic Diatessaron," Oriens 
Christianus, lxxxiii (1999), pp. 72-129. Recently, D. C. Parker, D. G. K. 
Taylor, and M. S. Goodacre have challenged the generally accepted view 
that 0212 is part of Tatian's Diatessaron (see their "The Dura-Europos 
Gospel Harmony," in Studies in the Early Text of the Gospels and Acts, 
ed. by D. G. K. Taylor [Atlanta, 19991, pp. 192-228). 


The History of New Testament 

Textual Criticism as Reflected 

in Printed Editions of the 

Greek Testament 


The Precritical Period 

The Origin and Dominance 
of the Textus Receptus 

I. From Ximenes and Erasmus 
to the Elzevirs 

Johannes Gutenberg's invention of printing by using movable type 
had the most momentous consequences for Western culture and 
civilization. Now copies of books could be reproduced more rapidly, 
more cheaply, and with a higher degree of accuracy than had ever 
been possible previously. Quite appropriately, the first major product 
of Gutenberg's press was a magnificent edition of the Bible. The text 
was Jerome's Latin Vulgate, and the volume was published at Mayence 
(Mainz) between 1450 and 1456. During the next 50 years, at least 100 
editions of the Latin Bible were issued by various printing houses. In 
1488, the first edition of the complete Hebrew Old Testament came 
from the Soncino press in Lombardy. Before 1500, Bibles had been 
printed in several of the principal vernacular languages of western 
Europe: Bohemian (Czech), French, German, and Italian. 

But, except for several short extracts, 1 the Greek New Testament 
had to wait until 1514 to come from the press. Why was there so 

1. In 1481, the Greek text of the hymns of Zachariah and Elizabeth 
(Luke 1) was printed at Milan in an appendix to a Greek Psalter. In 1504, 


138 The Text of the New Testament 

long a delay? Two reasons may be suggested that help to account for 
the lapse of some 60 years from Gutenberg's invention to the first 
printed Greek Testament. 

In the first place, the production of fonts of Greek type neces- 
sary for a book of any considerable size was both difficult and 
expensive. 2 The attempt was made to reproduce in print the appear- 
ance of minuscule Greek handwriting, with its numerous alternative 
forms of the same letter as well as its many combinations of two or 
more letters (ligatures)? Instead, therefore, of producing type for 
merely the 24 letters of the Greek alphabet, printers prepared about 
200 different characters. (Subsequently, these variant forms of the 
same letters were abandoned, until today there remain only the two 
forms of the lower-case sigma, a and g.) 

The principal cause that retarded publication of the Greek text 
of the New Testament was doubtless the prestige of Jerome's Latin 
Vulgate. Translations into the vernacular languages were not deroga- 
tory to the supremacy of the Latin text from which they were 
derived, but the publication of the Greek New Testament offered to 
any scholar acquainted with both languages a tool with which to crit- 
icize and correct the official Latin Bible of the Church. 

At length, however, in 1514, the first printed Greek New Testa- 
ment came from the press, as part of a polyglot Bible. Planned in 

the first six chapters of John in Greek (taken from a Gospel lectionary and 
retaining the rubrics) were printed at Venice in a book containing a Latin 
translation of the poems of Gregory Nazianzen. In 1514, John 1.1-14 in 
Greek was reprinted at Tubingen. 

2. The first book with continuous Greek type was a grammar of the 
Greek language entitled Erotemata, written by Constantine Lascaris and 
published at Milan in 1476; see Richard P. Breaden, "The First Book Printed 
in Greek," Bulletin of the New York Public Library, li (1947), pp. 586-92. 
Before 1476, several editions of Latin classics and Church fathers had ap- 
peared containing occasional Greek words and phrases. The honor of 
being the first to print Greek goes either to Johann Fust and Peter Schoeffer 
for their edition of Cicero's De officiis and Paradoxa, published at Mainz in 
1465, or to Conrad Sweynheym and Arnoldus Pannartz for their volume of 
Lactantius' Opera, issued at Subiaco, Italy, on 30 Oct. 1465. See Victor 
Scholderer, Greek Printing Types, 1465-1927 '(London, 1927), p. 1. 

3. For a list of many of the ligatures used in incunabula, see the Style 
Manual of the United States Government Printing Office (Washington, 
D.C., 1945), pp. 316-18, or Georg F. von Ostermann, Manual of Foreign 
Languages (New York, 1952), pp. 105-8. 

The Precritical Period 139 

1502 by the cardinal primate of Spain, Francisco Ximenes de Cisneros 
(1436-1517), this magnificent edition of the Hebrew, Aramaic, Greek, 
and Latin texts was printed at the university town of Alcala 4 (called 
Complutum in Latin). 5 Known as the Complutensian Polyglot, 6 the 
project was under the editorial care of several scholars, of whom 
Diego Lopez de Zuniga (Stunica) is perhaps the best known. 7 Vol- 
ume v, containing the New Testament and a Greek glossary with Latin 
equivalents, was printed first, its colophon bearing the date 10 January 
1514. Volume vi, the appendix, containing a Hebrew lexicon and an 
elementary Hebrew grammar, was printed next, in 1515. The four 
volumes of the Old Testament appeared last, the colophon of volume 
iv bearing the date 10 July 1517. The sanction of Pope Leo X, printed 
in volume i, was not obtained until 22 March 1520; but it appears that 
for some reason the Polyglot was not actually circulated (i.e., pub- 
lished) until about 1522. 

The four volumes that contain the Old Testament present the 
Hebrew text, the Latin Vulgate, and the Greek Septuagint in three 
columns side by side on each page, together with the Aramaic Targum 

4. See Julian Martin Abad, "The Printing Press at Alcala de Henares, 
The Complutensian Polyglot Bible," in The Bible as Book, The First Printed 
Editions, ed. by Kimberly Van Kampen and Paul Saenger (London, 1999), 
pp. 101-18. 

5. See Jerry H. Bentley, Humanists and Holy Writ: New Testament 
Scholarship in the Renaissance (Princeton, 1983), pp. 70-111; B. M. 
Metzger, "History of Editing the Greek New Testament," Proceedings of the 
American Philosophical Society, cxxxi (1987), pp. 148-58; M. V. Spotorno, 
"The Textual Significance of the Spanish Polyglot Bibles," Sepharad, Ixii 
(2002), pp. 375-92. 

6. Of the original 600 sets that were printed, the locations of 97 are 
known today; see James P. R. Lyell, Cardinal Ximenes (London, 1917); 
Mariano Revilla Rico, La Poliglota de Alcala, estudio historico-critico 
(Madrid, 1917). 

7. One of the lesser known collaborators in the project, who never- 
theless played an important role in the preparation of the Greek text of the 
Septuagint and of the New Testament for the Polyglot, was Demetrius 
Ducas, a Greek from Crete whom Ximenes brought to Alcala to teach in 
his academy and to edit Greek books; see the account concerning Ducas 
in Deno J. Geanakoplos, Greek Scholars in Venice: Studies in the Dissemi- 
nation of Greek Learning from Byzantium to Western Europe (Cambridge, 
MA, 1962), pp. 238 ff. 

140 The Text of the New Testament 

of Onkelos for the Pentateuch at the foot of the page, accompanied 
by a Latin translation. The Greek type used in the New Testament 
volume is modeled after the style of the handwriting in manuscripts 
of about the eleventh or twelfth century and is very bold and elegant 
(Fig. 22). 8 It is printed without rough or smooth breathing marks and 
accented according to a system never heard of before or since: 
monosyllables have no accent, while the tone syllable in other words 
is marked with a simple apex, resembling the Greek acute accent 
mark. Each word or group of Greek words is coded to the adjacent 
column of the Latin Vulgate by small supralinear roman letters, thus 
assisting readers with little Greek to find the equivalent words 
in each column. The Septuagint is printed with the familiar cursive 
style of Greek characters popularized by Aldus Manutius, the famous 
Venetian printer. 

What Greek manuscripts lie behind the text of the Compluten- 
sian New Testament has never been satisfactorily ascertained. 9 In his 
dedication to Pope Leo X, after mentioning the pains that he had 
taken to secure Latin, Greek, and Hebrew manuscripts, Ximenes 
continues: "For Greek copies indeed we are indebted to your 
Holiness, who sent us most kindly from the Apostolic Library very 

8. According to Scholderer, the unknown designer of the Greek 
font of type used in the Complutensian New Testament incorporated 
certain features of Nicolas Jenson's Greek font, which was put into use 
in 1471. 

In fact, it [the Greek of the Complutensian New Testament] is the last and most 
beautiful example of the Jensonian class of type, carrying it to the limit of its 
possibilities, and is fairly entitled to its generally acknowledged position as the 
king of all Greek fonts, although its full perfection was not disclosed until the 
present century, when Robert Proctor completed a revival of the lower-case in 
an enlarged copy by adding a full set of capitals adapted to it. The Greek of 
tiie Complutensian New Testament remains the only original contribution of 
Spain to Hellenic typography. (Scholderer, op. cit., p. 10) 

9. See, e.g., Franz Uelitzsch, Studien zur Entstehungsgeschichte der 
Polyglottenbibel des Cardinals Ximenes (Programm; Leipzig, 1871; also 
published [in German] at London under the title Studies on the Compluten- 
sian Polyglott [1872]); Samuel Berger, La Bible au seizieme siecle (Paris, 
1879), pp. 51-8; Delitzsch, Fortgesetzte Studien zur Entstehungsgeschichte 
der Complutensischen Polyglottenbibel (Programm, Leipzig, 1886); 

M. Goguel, "Le texte et les editions du Nouvcau Testament grec," Revue 
de I'histoire des religions, lxxxii (1920), pp. 14-18. 

The Precritical Period 




«bc , owK'iAOK<iia«JV'" Io l , '**«l''*'<«'P*«V'<mY l^"p»taii«rtM)«*ibeKtfftlB*lM 
■tet^irapiA«M(^oirro4/c^*t<$<'CK^Gao*a«o|i •iwduWiBoe'iiciiii'Mi'i «« 

|«^*«K4(ni , oAii({a,'nioputia,«iDO}<Hp<a,'TtAto pkwe"oinniqwBm^iwJWK I fcB»c«to 

'AdXou.'watOHWfcK;. 4^»yp"JTa<,'i!crTaXd wiadl|e:'<i5(tiiBo«ic' t <o l o . '^ iinliB ii M l c i 


*npdocrovoi. £•$.*. 

■id'ayamXtSynro^' «*» '<Sj»p»m 'rraq/o 
avr6|i # KerraKp{|iiic. /Tta ' yap* avrd ' npaer * 

cmi^Kp{y«)i. * oi&autyAs'oTi /"To* Kpi'ua "Tov 

»• i^iio^owolo'ioia^Ta^nrpskrcqitac "koo 
toOKoifavrd ,'6-n'av'(i«psufH(7o'iep{ua/1ou 


Tov , Kov'lM<;"a(x>XH<; 'itawlH<*4KKpo»v.uiac*Ka 
Ta9po(jEi(;.faYVoci>v 'owlo'XpHcrTojirlow "#i«« 

pdTHTorOO^KOB'auCTQJIOHTOll'KapiiQll '*H0<XU 

p(J«C * «wnrr« 'op-y^sy HJuipofopym *koi *ct 
iioKoX\H'MK , Kai'2ktKOiOKpi<ri(»;/'7ow , «tov, 

•oC*on»ia>«i , SKaffT«»'KaTttla«spya*auT«u, -q«*tt(fckrTnia«»1'ta»du}'op»*i 
•lon'Jiiji'KafvmoxiOimiJ'tpyou'ayaWu'ao Ittr.Mf^'ftjJspanftii'booi'oBW'fi 
|(^ , Kai'TiiiH^Kai'apitopiJiap'nnjHTo\>ffi,'Ja>> 

w'a.'svxio^KCB'opyH ' »Ai4/ii; 'koi * tmpoxm == 
pia »«iTi''waea|j*\|A/XH(f'QiwpcbfToy / "fov'Ka* 3 
T^>yo3oj«|iow/lo , Kep«j^*ioviaio\^Ii^fp<»TO|» 
'KeB*iXXHtio<.*i6fa*JM*Kai'"riaH'Kai 'spHjiK 
r tt*«Bp»Tou*K<B , tA>*<|ii .*oy 'yup'son'OTpo* 
o<BTioAH^ia*05aptt"T»'«tto . *<woi ' ycp'apd* 
iiow;"KUQpTop, 'ay&uaHfKafianoXovyrrau .'kct 
•6ooi'tp'p6juai'HiiapTo^.'2ud , p6uov*Kpi««oiJ 
■roi.'oy'yap.a'aKpocrrai * Tow* jkJuov 'Aiicatoi 
•wapdlco»»Kb,'aXA«>i * ojoihtqi it lov' pduov 
j«3|i*cXO|rro'9vaii > lo'7ov > |iduou*asoti%;*ov^ 
tot'pduo|i k w)"ixo)Rc<;^avToiCEi0i'pduo<:>'d 
T^tct|i2iiK|(u|(T<a/7o'^>yo|«lQV > |i6iiov 'ypa 
nroi»"iiJ(Tan;"KapAiaK;'awTwiJ.'ffvuiKxpTu = 
pov«M<?avToe|J 1m; '<JvytiAH<rEa><; , • koi ' jwto - 
|v*aXXHAo5(j/l<op ' Aoyiojutojj ' KaTHyopovp* 

Figure 22 Complutensian Polyglot Bible, vol. v, first printed Greek New 
Testament (1514), Rom. 1.27-2.15 (see pp. 138-40). Actual size 14 X 9'A 

«umd>ofo6:'fugtx»."clat(»: > fci& w(! ) l i i> 

8 ni«fflr'mottt:l>a i foUl<)'«f»n6B*W , ft 
qni'ioafcniiiiM'fitdcnAw. £#.1. 

dioe'topfiim'cWdiio.' Eadan'aifa 
aum , t>ci"dl*laJm'T«uiMon , in'««'<i 
• tt lu' as at. , eaftimM*aflrto<'o ' bono 
-qui-ujdiu8'coe*qurialu*afiu« f rtco 

'oaf'Bn'Nwniw'boiiiuoe'diKJ oo» 
'cf MnflittcVtHofisanimittmVoflt&itf 
1 Jgnwas'quonwm 'bcraemoe-oo oooo 
'ribi'iram'in'oif'irtWrtudaooflte oo 


■mAm*tu<ki t p»muinoca>099as»xioo 

'omm'epmm'lxmmi'ioito <nsoss» 
•pjfaiPf gww.'THd'd'tmatftprtolBfe.' *" 
no^iipud'ocfl.'iautc&B'ad-ftiic-ira* £«. • 
•p«aiiCTflK'flB<rk0fpertbu«.*ft«e *£» 

buHe.'fJon'cni'aiuliivvWItiguj'wflifUBf *■•* 


■m , <mW^twi>«:%itinK«tiifl ,, twfcKK*' 
'ctwiacmu «ipfojum:*o'inta- ««» 
*fc* mkxm ' coflttaaommi T iaiifjnmim 

142 The Text of the New Testament 

ancient codices, both of the Old and the New Testament; which have 
aided us very much in this undertaking." 10 

Though the Complutensian text was the first Greek New Testa- 
ment to be printed, the first Greek New Testament to be published 
(i.e., put on the market) was the edition prepared by the famous 
Dutch scholar and humanist Desiderius Erasmus of Rotterdam 
(1469-1 536). n It cannot be determined exactly when Erasmus first 
decided to prepare an edition of the Greek Testament, but on a visit 
to Basle in August 1514 he discussed (probably not for the first time) 
the possibility of such a volume with the well-known publisher 
Johann Froben. Their negotiations seem to have broken off for a 
time but were resumed in April 1515 while Erasmus was on a visit at 
the University of Cambridge. It was then that Froben importuned 
him through a mutual friend, Beatus Rhenanus, to undertake imme- 
diately an edition of the New Testament. Doubtless, Froben had 
heard of the forthcoming Spanish polyglot Bible and, sensing that 
the market was ready for an edition of the Greek New Testament, 
wished to capitalize upon that demand before Ximenes' work would 
be finished and authorized for publication. Froben's proposal, which 
was accompanied by a promise to pay Erasmus as much as anyone 
else might offer for such a job (se daturum pollicetur, quantum alius 
quisquam), apparently came at an opportune moment. Going to 
Basle again in July of 1515, Erasmus hoped to find Greek manuscripts 
sufficiently good to be sent to the printer as copy to be set up in type 
along with a Vulgate text that "Erasmus had rather extensively revised, 
so as to bring it in line with the Greek text, printed opposite the 
Vulgate in parallel columns." 12 To his vexation, the only manuscripts 

10. Some have doubted the truth of this statement, for Leo had been 
elected pope less than a year before the New Testament volume was fin- 
ished and therefore (so it is argued) could hardly have sent manuscripts in 
time to be useful (see Marvin R. Vincent, A History of the Textual Criticism 
of the New Testament [New York, 18991, p. 50). One should observe, how- 
ever, that Ximenes does not explicitly say that the manuscripts were sent 
during Leo's pontificate; it is altogether possible, as Hug suggested long 
ago, that they had been sent during the pontificate of the previous pope, 
Julius II, through the intervention of Cardinal de Medici, who had great 
influence over Julius and who in turn succeeded him to the papal throne 
(J. L. Hug, Einleitung in die Schriften des Neuen Testaments [Tubingen, 
1826], § 55). 

11. SeeJ. H. Bentley, Humanists and Holy Writ, pp. 112-93. 

12. The quotation is from Bentley, op. cit., p. 135. 

The Precritical Period 143 

available on the spur of the moment required a certain amount of 
correcting before they could be used as printer's copy. 13 

The printing began on 2 October 1515, and in a remarkably short 
time (1 March 1516), the entire edition was finished, a large folio vol- 
ume of about 1,000 pages that, as Erasmus himself declared later, 
was "precipitated rather than edited" ipraecipitatum verius quam 
editurri). Owing to the haste in production, the volume contains hun- 
dreds of typographical errors; in fact, Scrivener once declared, "[It] is 
in that respect the most faulty book I know." 14 Since Erasmus could 
not find a manuscript that contained the entire Greek Testament, he 
utilized several for various parts of the New Testament. For most of 
the text he relied on two rather inferior manuscripts from a monastic 
library at Basle, one of the Gospels (Fig. 23) and one of the Acts and 
Epistles, both dating from about the twelfth century. 15 Erasmus com- 
pared them with two or three others of the same books and entered 
occasional corrections for the printer in the margins or between the 
lines of the Greek script. l6 For the Book of Revelation, he had but 

13. Most of the manuscripts that Erasmus used in the preparation of 
his editions of the New Testament came from the collection that had been 
bequeathed in 1443 to the Dominican monastery at Basle by John 
Stojkovic of Ragusa, one of the cardinals created by the anti-pope, Felix V; 
see Aloysius Krchnak, De vita et operibus Ioannis de Ragusia (Lateranum, 
n.s., xxvi; Rome, I960), pp. 2-3; R. W. Hunt, "Greek Manuscripts in the 
Bodleian Library from the Collection of John Stojkovic of Ragusa," in 
Studia Patristica, vol. vii, part 1, ed. by F. L. Cross (Texte und Unter- 
suchungen, xcii; Berlin, 1966), pp. 75-82. See also Bo Reicke, "Erasmus 
und die neutestamentliche Textgeschichte," Theologische Zeitschrift, xxii 
(1966), pp. 254-65, who discusses the several Greek manuscripts, now in 
the university library at Basle, used by Erasmus for his first edition of the 
Greek New Testament. See most recently P.-Y. Brandt, "Manuscrits grecs 
utilises par Erasme pour son edition du Novum Instrumentum de 1516," 
Theologische Zeitschrift, liv(1998), pp. 120-4. 

14. F. H. A. Scrivener, A Plain Introduction to the Criticism of the New 
Testament, 4th ed., ii (London, 1894), p. 185. 

15. Though some have dated the manuscript of the Gospels in the 
fifteenth century (so Scrivener [doubtfully!, Kenyon, and von Dobschiitz), 
Gregory, Eberhard Nestle, von Soden, and Clark have assigned it to the 

16. These corrections are described by Kenneth W. Clark, "The 
Erasmian Notes in Codex 2," in Studia Evangelica, ed. by K. Aland, F. L. 
Cross, et al. (Texte und Untersuchungen, lxxiii; Berlin, 1959), pp. 749-56. 
See also C. C. Tarelli, "Erasmus's Manuscripts of the Gospels," Journal of 
Theological Studies, xliv (1943), pp. 155-62. 

l *$E 

is- ■*: %./ * * » w » .** f«i^W . 

-j" Ortftt^V* -r«l4^<MC^«mi{-n'<rvn • J Niuji 

f ^y^W"** H*"^ «KA*«ii»4*»|i ••• «*t+sp •• 

JML* j • ' •' "A' \ • , r »."•'"■* IB 


Figure 23 Greek Gospel MS. 2 (twelfth century). University Library, Basle; 
Luke 6 20-30; one of the inferior manuscripts used by Erasmus for his first 
edition of the Greek New Testament (see pp. 142-3), with his corrections 
and annotations for the printer (e.g., in lines 8 and 9, an oblique stroke sep- 
arates the definite article from the following word, and in the lower margin 
is Erasmus' addition of JtQooevxeciOe vjieq rmv ejrrfQea^ovzcov vfidg, which 
the scribe had accidentally omitted from the text of verse 28, third line from 
the bottom). Actual size 7K X 6 inches. 


The Precritical Period 145 

one manuscript, dating from the twelfth century, which he had bor- 
rowed from his friend Reuchlin. Unfortunately, this manuscript lacked 
the final leaf, which contained the last six verses of the book. Instead 
of delaying the publication of his edition while trying to locate an- 
other copy of Revelation in Greek, Erasmus (perhaps at the urging of 
his printer) depended on the Latin Vulgate and translated the miss- 
ing verses into Greek. As would be expected from such a procedure, 
here and there in Erasmus' self-made Greek text are readings that 
have never been found in any known Greek manuscript of these 
verses — but that are still perpetuated today in printings of the so- 
called Textus Receptus of the Greek New Testament. 17 

Even in other parts of the New Testament Erasmus occasionally 
introduced into his Greek text material taken from the Latin Vulgate. 
Thus, in Acts 9-6, the question that Paul asks at the time of his con- 
version on the Damascus road, "And he trembling and astonished 
said, Lord, what will you have me to do?," was frankly interpolated 
by Erasmus from the Latin Vulgate. This addition, which is found in 
no Greek manuscript at this passage, became part of the Textus 
Receptus, from which the King James Version was made in 1611. 

The reception accorded Erasmus' edition, the first published 
Greek New Testament, was mixed. On the one hand, it found many 
purchasers throughout Europe. Within 3 years a second edition was 
called for, and the total number of copies of the 1516 and 1519 edi- 
tions amounted to 3,300. The second edition became the basis of 
Luther's German translation. 18 On the other hand, in certain circles, 
Erasmus' work was received with suspicion and even outright hos- 
tility. The replacement in Erasmus' second edition of Jerome's Latin 
Vulgate with Erasmus' own more elegant Latin translation, which dif- 
fered in many respects from the wording of the Vulgate, was regarded 
as a presumptuous innovation. Particularly objectionable were the 

17. For example ogOgivog (22.16); iXOe twice, D.Oetco (22.17); 
ovftfiagTVQOVfiai ydg . . . emxidfj ngograma (22.18); cupcugfj 
pifiXov . . . CKpaigijoei (future for d^eAet!!), ftifiXov (second occurrence) 
(22.19); vpi&v (22.21). 

18. It has often been debated how far Luther's translation rests on the 
Greek text. H. Dibbelt (Arcbivfur Reformationsgeschichte, xxxviii [1941], 
pp. 300-30) maintained that the translation reflects only an occasional 
consultation of the Greek; II. Bornkamm {Theologische Literaturzeitung, 
lxxii [1947], pp. 23-8) held that Luther translated from the combination of 
Greek and Latin texts in Erasmus' edition and from the Vulgate, which he 
had in his head. 

146 The Text of the New Testament 

brief annotations in which Erasmus sought to justify his translation. 
He included among the philological notes not a few caustic com- 
ments aimed at the corrupt lives of many of the priests. In the words 
of J. A. Froude, "The clergy's skins were tender from long impunity. 
They shrieked from pulpit and platform, and made Europe ring with 
their clamour." 19 As a result, "universities, Cambridge and Oxford 
among them, forbade students to read Erasmus' writings or book- 
sellers to sell them." 20 

Among the criticisms leveled at Erasmus, the most serious 
appeared to be the charge of Stunica, one of the editors of Ximenes' 
Complutensian Polyglot, that his text lacked part of the final chapter 
of 1 John, namely the Trinitarian statement concerning "the Father, 
the Word, and the Holy Ghost: and these three are one. And there are 
three that bear witness in earth" (1 John 5.7-8, King James Version). 
Erasmus replied that he had not found any Greek manuscript that 
contained these words, though he had in the meanwhile examined 
several others besides those on which he relied when first preparing 
his text. In an unguarded moment, Erasmus may have promised that 
he would insert the Comma Johanneum, as it is called, 21 in future 
editions if a single Greek manuscript could be found that contained 
the passage. 22 At length, such a copy was found — or was made to 
order! As it now appears, the Greek manuscript had probably been 
written in Oxford about 1520 by a Franciscan friar named Froy (or 
Roy), who took the disputed words from the Latin Vulgate. 23 Erasmus 
inserted the passage in his third edition (1522), but in a lengthy foot- 
note that was included in his volume of annotations, he intimated his 

19. J- A. Froude, Life and Letters of Erasmus (New York, 1896), p. 127. 
See also August Bludau, Die beiden ersten Erasmus-Ausgaben des Neuen 
Testaments und ihre Gegner (Freiburg, 1902). 

20. Froude, op. cit., p. 138. 

21. The word comma in this usage means a short clause of a 

22. It should, however, be noted that Henk Jan de Jonge, a specialist 
in Erasmian studies, could find no explicit evidence that supports this 
frequently made assertion concerning a specific promise made by Erasmus; 
see his "Erasmus and the Comma Johanneum" Ephemerides Theologicae 
Lovanienses, Ivi (1980), pp. 381-9. 

23. See J. Rendel Harris, The Origin of the Leicester Codex of the New 
Testament (London, 1887), pp. 40-53; C. H. Turner, The Early Printed 
Editions of the Greek Testament (Oxford, 1924), pp. 23-4. 

The Precritical Period 147 

suspicion that the manuscript had been prepared expressly in order 
to confute him. 24 

Among the thousands of Greek manuscripts of the New Testa- 
ment examined since the time of Erasmus, only eight are known to 
contain this passage. In four of the eight, the Comma appears in the 
text; in the other four, it is a marginal addition serving as an alterna- 
tive or variant reading. The eight are the following, listed according 
to the Gregory- Aland enumeration: 25 

61: the Codex Montfortianus, an early sixteenth-century manuscript at 

Trinity College, Dublin. This codex was copied from a tenth-century 

manuscript at Lincoln College, Oxford, that did not have the 

Comma. Insertions elsewhere in the Montfortianus copy have been 

retro verted from the Latin. 
88 vr : a variant reading in a sixteenth-century hand, added to the 

twelfth-century Codex Regius at Naples. 
221 vr : a variant reading added to a tenth-century manuscript in the 

Bodleian Library at Oxford. 
429 vr : a variant reading added to a fifteenth-century manuscript at 

629: the Codex Ottobonianus at the Vatican. It is of the fourteenth 

century and has a Latin text alongside the Greek, which has been 

revised according to the Vulgate. 
636 vx : a variant reading added to a fifteenth-century manuscript at 

918: a sixteenth-century manuscript at the Escorial, Spain. 
2318: an eighteenth-century manuscript influenced by the Clementine 

Vulgate, at Bucharest, Rumania. 

The oldest known citation of the Comma is in a fourth-century 
Latin treatise entitled Liber apologeticus (Chapter 4), attributed either 

24. Today, the codex (designated manuscript 61 in the Gregory-Aland 
listing), which is in the library of Trinity College, Dublin, opens almost of 
its own accord at 1 John 5 — so often has it been consulted at this passage! 

25- Omitted from the list is Codex Ravianus (Tischendorf at 110), pre- 
served in the Royal Library at Berlin. It is of the sixteenth century and had 
been copied from the printed Complutensian Polyglot Greek text. For 
further discussion of the spurious nature of the Comma, see R. E. Brown, 
The Epistles of John (Garden City, NJ, 1982), pp. 775-87; B. M. Metzger, 
A Textual Commentary on the Greek New Testament, 2nd ed. (New York, 
1994), pp. 647-9. 

148 The Text of the New Testament 

to Priscillian or to his follower, Bishop Instantius of Spain. The 
Comma probably originated as a piece of allegorical exegesis of 
the three witnesses and may have been written as a marginal gloss in 
a Latin manuscript of 1 John, whence it was taken into the text of the 
Old Latin Bible during the fifth century. The passage does not appear 
in manuscripts of the Latin Vulgate before about A.D. 800. In view of 
its inclusion in the Clementine edition of the Latin Vulgate (1592), in 
1897 the Holy Office in Rome, a high ecclesiastical congregation, 
made an authoritative pronouncement, approved and confirmed by 
Pope Leo XIII, that it is not safe to deny that this verse is an authen- 
tic part of St. John's Epistle. 26 Modern Roman Catholic scholars, 
however, recognize that the words do not belong in the Greek 
Testament; for example, the four bilingual editions of the New Tes- 
tament that were edited by Bover, Merk, Nolli, and Vogels include 
the words as part of the Vulgate text approved by the Council of 
Trent but reject them from the Greek text that faces the Latin on the 
opposite page. 27 

Subsequently, Erasmus issued a fourth and definitive edition 
(1527), which contains the text of the New Testament in three paral- 
lel columns, the Greek, Erasmus' own Latin version, and the Latin 
Vulgate. He had seen Ximenes' Polyglot Bible shortly after the pub- 
lication of his own third edition in 1522 and wisely decided to avail 
himself of its generally superior Greek text to the improvement of his 
own. In the Book of Revelation, for example, he altered his fourth 
edition in about 90 passages on the basis of the Complutensian text. 
A fifth edition, which appeared in 1535, discarded the Latin Vulgate 
but differed very little from the fourth regarding the Greek text. 

Thus, the text of Erasmus' Greek New Testament rests upon a 
half-dozen minuscule Greek manuscripts. The oldest and best of 
these manuscripts (Codex 1, a minuscule of the tenth century, which 
agrees often with the earlier majuscule text) he used least because he 

26. The Holy Office declared subsequently (2 June 1927) that its 
decree was not intended to hinder Catholic scholars from thoroughly 
investigating the matter and from espousing an opinion contrary to the 
authenticity of the passage, provided that they profess themselves ready 
to stand by the judgment of the Church; see H. Denzinger and K. Rahner, 
Enchiridion symbolorum, 28th ed. (Freiburg, 1952), n. 2198; idem, 
Enchiridion biblicum, 3rd ed. (Rome, 1956), n. 136. 

27. For a full discussion by a noted Roman Catholic textual scholar, see 
Teofilo Ayuso Marazuela, "Nuevo estudio sobre el 'Comma loanneum,'" 
Biblica, xxviii (1947), pp. 83-112, 216-35; xxix (1948), pp. 52-76. 

The Precritical Period 149 

was afraid of its supposedly erratic text! Erasmus' text is inferior in 
critical value to the Complutensian, yet because it was the first on the 
market and was available in a cheaper and more convenient format, 
it attained a much wider circulation and exercised a far greater influ- 
ence than its rival, which had been in preparation from 1502 to 1514. 
In addition to Erasmus' five editions mentioned above, more than 
30 unauthorized reprints are said to have appeared at Venice, 
Strasbourg, Basle, Paris, and other places. 

Subsequent editors, though making a number of alterations in 
Erasmus' text, essentially reproduced this debased form of the Greek 
Testament. Having secured an undeserved preeminence, what came 
to be called the Textus Receptus of the New Testament resisted for 
400 years all scholarly efforts to displace it in favor of an earlier and 
more accurate text. The highlights of this history are as follows. 

The first edition of the whole Bible in Greek was published in 
three parts in February 1518 at Venice by the celebrated Aldine Press. 
The New Testament, which is dedicated to Erasmus, follows the first 
edition of Erasmus so closely as to reproduce many typographical 
errors — even those that Erasmus had corrected in a list of errata! 

A beautifully printed pocket-sized edition (its pages measure 
3 by 4 inches) was produced in two volumes (6l6 pp., 475 pp.) by 
loannes Antonius de Nicolinis de Sabio at Venice in 1538. Its text, 
edited by Melchiorre Sessa, is curiously eclectic, depending now on 
Erasmus, now on the Aldine text, and occasionally departing from all 
previous editions. 28 Like several other early editions, it contains in 
Greek certain "helps for readers," such as lists of chapter headings, 
lives of the Evangelists, hypotheses (introductions) to the several 
books, and accounts of the journeys of Paul and of his martyrdom. 29 

The famous Parisian printer and publisher Robert Estienne, 
latinized as Stephanus (1503-59), issued four editions of the Greek 
Testament, three at Paris (1546, 1549, and 1550) and the last 
at Geneva (1551), where he spent his final years as a professed 

28. For an analysis of the textual affinities of this extremely rare edi- 
tion, see W. H. P. Hatch, "An Early Edition of the New Testament in 
Greek," Harvard Theological Review, xxxiv (1941), pp. 69-78. The rarity of 
copies of the edition may be gauged from the statement of Reuss that after 
diligent search he was unable to locate a copy anywhere in the libraries of 
Germany. Hatch indicated that only seven copies of the complete edition 
and one copy of volume ii were known to exist. 

29- These "helps for readers" were drawn from minuscule manuscripts 
of the New Testament; see pp. 39-40. 

1 50 The Text of the New Testament 

Protestant. 30 The three Parisian editions are most sumptuously 
printed, with type cast at the expense of the French government. The 
handsome third edition, of folio size (8% by 13 inches), is the first 
Greek Testament that has a critical apparatus; Stephanus entered on 
the inner margins of the pages variant readings from 14 Greek 
codices as well as many readings from the Complutensian Polyglot. 
One of the manuscripts that he cited is the famous Codex Bezae, 
which had been collated for him, he says, "by friends in Italy." 

The text of Stephanus' editions of 1546 and 1549 was a com- 
pound of the Complutensian and Erasmian editions; the third edition 
(1550) approaches more closely the text of Erasmus' fourth and fifth 
editions. As it happened, Stephanus' third edition became for many 
persons, especially in England, the received or standard text of the 
Greek Testament. 

Stephanus' fourth edition (1551), which contains two Latin ver- 
sions (the Vulgate and that of Erasmus) printed on either side of the 
Greek text, is noteworthy because in it for the first time the text was 
divided into numbered verses. It has often been stated that 
Stephanus marked the verse divisions while journeying "on horse- 
back" and that some of the infelicitous divisions arose from the 
jogging of the horse that bumped his pen into the wrong places. 31 
Stephanus' son does indeed assert that his father did the work while 
on a journey {inter equitandum) from Paris to Lyons, but the most 
natural inference is that the task was accomplished while resting at 
the inns along the road. 32 

30. See Elizabeth Armstrong, Robert Estienne, Royal Printer: An 
Historical Study of the Elder Stephanus (Cambridge, 1954), pp. 211 ff. 

31. So, e.g., A. T. Robertson, An Introduction to the Textual Criticism 
of the New Testament, 2nd ed. (New York, 1928), p. 100. 

32. For wide-ranging information concerning verse division in the New 
Testament, see William Wright's article "Verse" in the first edition (the article 
is shortened in later editions) of John Kino's Cyclopcedia of Biblical Litera- 
ture, ii (London, 1845; New York, 1846), pp. 905-14, supplemented by 
Wright's articles in The Christian Remembrancer, n.s., iii (January-June 
1842), pp. 455-69, 672-90. For a list of differences in verse division among 
about 50 editions of the Greek New Testament, see Ezra Abbot's excursus in 
Caspar Rene Gregory's Prolegomena, being vol. iii of Tischendorf s Novum 
Testamentum Graece, 8th ed., part 1 (Leipzig, 1884), pp. 167-82, reprinted in 
English in Abbot's posthumously published volume entitled The Authorship 
of the Fourth Gospel and Other Critical Essays (Boston, 1888), pp. 464-77. 

The Precritical Period 151 

In 1553, Stephanus' folio edition of 1550 was reprinted in a small 
volume {3% by 5% inches) by Jean Crispin (or Crespin), the French 
printer of Geneva, who published many editions of the Scriptures in 
various languages, including the second quarto English Geneva 
Bible of 1570. Crispin reproduced the text of Stephanus with only 
half a dozen minor alterations. The variant readings of the 1550 folio 
edition are also reproduced, 33 though without Stephanus' sigla refer- 
ring to individual manuscripts. It was either Stephanus' folio edition 
or Crispin's pocket-sized reprint that William Whittingham and his 
fellow Protestant refugees from England utilized when they prepared 
their English translation of the New Testament (Geneva, 1557), the 
first English version to include variant readings in the margins. 34 

Theodore de Beze (Beza, 1519-1605), the friend and successor 
of Calvin at Geneva and an eminent classical and biblical scholar, 
published no fewer than nine editions of the Greek Testament be- 
tween 1565 and 1604, and a tenth edition appeared posthumously in 
1611. Only four of them, however, are independent editions (those 
of 1565, 1582, 1588-9, and 1598), the others being smaller reprints. 
Accompanied by annotations and his own Latin version, as well as 
Jerome's Latin Vulgate, these editions contain a certain amount of 
textual information drawn from several Greek manuscripts that Beza 
had collated himself, as well as the Greek manuscripts collated by 
Henry Stephanus, son of Robert Stephanus. Noteworthy among 
Beza's own manuscript possessions were Codex Bezae and Codex 
Claromontanus, though he made relatively little use of them, for they 
deviated too far from the generally received text of the time. Beza 
seems also to have been the first scholar to collate the Syriac New- 
Testament, which was published in 1569 by Emmanuel Tremelltus. 
For Acts and 1 and 2 Corinthians he utilized information from the 
Arabic version put at his disposal by Franciscus Junius. Despite the 
variety of this additional textual evidence available to Beza, which is 
reflected chiefly in his annotations, the Greek text he printed differs 
little from Stephanus' fourth edition of 1551. The importance 
of Beza's work lies in the extent to which his editions tended to 

33. The variant reading at Luke 17.35, however, is introduced into 
Crispin's text itself. 

34. The translation, with minor alterations, and the variant readings 
were incorporated in the Geneva Bible of 1560; see the discussion of the 
variant readings in B. M. Metzger, "The Influence of Codex Bezae upon the 
Geneva Bible of 1560," New Testament Studies, viii (196l), pp. 72-7. 

152 The Text of the New Testament 

popularize and to stereotype the Textus Receptus. The King James 
translators of 1611 made large use of Beza's editions of 1588-9 and 

In 1624, Bonaventure and Abraham Elzevir, two enterprising 
printers at Leiden, 35 published a small and convenient edition of the 
Greek Testament, the text of which was taken mainly from Beza's 
smaller 1565 edition. The preface to the second edition, which ap- 
peared in 1633, makes the boast that "[the reader has] the text now 
received by all, in which we give nothing changed or corrupted." 36 
Thus, from what was a more or less casual phrase advertising the 
edition (what modern publishers might call a "blurb"), there arose 
the designation "Textus Receptus," or commonly received, standard 
text. Partly because of this catchword, the form of the Greek text in- 
corporated in the editions that Stephanus, Beza, and the Elzevirs 
published succeeded in establishing itself as "the only true text" of 
the New Testament and was slavishly reprinted in hundreds of sub- 
sequent editions. It lies at the basis of the King James Version and of 
all the principal Protestant translations in the languages of Europe 
prior to 1881. So superstitious has been the reverence accorded the 
Textus Receptus that in some cases attempts to criticize or emend it 
have been regarded as akin to sacrilege. Yet, its textual basis is 
essentially a handful of late and haphazardly collected minuscule 
manuscripts, and in a dozen passages its rendering is supported by 
no known Greek witness. 

II. The Collection of Variant Readings 

The next stage in the history of New Testament textual criticism 
is characterized by assiduous efforts to assemble variant readings 

35. The printing house of Elzevir (properly Elzevier), founded by 
Louis Elzevier (1540-1617), issued many beautiful editions of classical 
authors from 1595 to 1681. On the printing history of the Elzevir family, 
see D. W. Davies, The World of the Elzevirs, 1580- 1712 (The Hague, 
1954); P. R. Sellin, Daniel Heinsius and Stuart England (London, 1968). 

36. "Textum ergo habes, nunc ab omnibus receptum: in quo nihil 
immutatum aut corruptum damus." The preface to the second edition was 
written by Daniel Heinsius (1580-1655) and the editor was Jeremias 
Hoelzlin (1583-1641); both were professors at Leiden. For further informa- 
tion, see H. J. de Jonge in Miscellanea Neotestamentica, ed. by Tj. Baarda, 
A. F. J. Klijn, and W. C. van Unnik, i (Leiden, 1978), pp. 105-28. 

The Precritical Period 153 

from Greek manuscripts, versions, and fathers. For almost two cen- 
turies scholars ransacked libraries and museums, in Europe as well 
as the Near East, for witnesses to the text of the New Testament. But 
almost all of the editors of the New Testament during this period were 
content to reprint the time-honored but corrupt Textus Receptus, 
relegating the evidence for the earlier readings to the apparatus. An 
occasional brave soul who ventured to print a different form of 
Greek text was either condemned or ignored. 

The first systematic collection of variant readings (those given in 
the margin of Stephanus' 1550 edition had been assembled some- 
what at random) was included in the Polyglot Bible edited by Brian 
Walton (l60O-6l) and published at London in 1655-7 in six folio 
volumes. 37 The fifth volume (1657) contains the New Testament in 
Greek, Latin (both the Vulgate and the version of Arius Montanus), 
Syriac, Ethiopic, Arabic, and (for the Gospels) Persian. The Greek 
text as well as each of the Eastern versions is supplied with a literal 
translation into Latin. The Greek text is that of Stephanus' 1550 
edition, with slight alterations. At the foot of the page are variant 
readings from Codex Alexandrinus, which had recently been pre- 
sented (1627) by Cyril Lucar, the patriarch of Constantinople, to 
Charles I. In the sixth volume of the Polyglot, the appendix, Walton 
included a critical apparatus, prepared by Archbishop Ussher, of 
variant readings derived from 15 other authorities, to which were 
added the variants from Stephanus' margin. 38 

In 1675, Dr. John Fell (1625-86), 39 dean of Christ Church and 
afterward bishop of Oxford, issued anonymously a small volume 
(3% by 6Vi inches), the first Greek Testament to be published at 

37. See P. N. Miller, "The 'Antiquarianization' of Biblical Scholarship 
and the London Polyglot Bible (1653-57)," Journal of the History of Ideas, 
lxii (2001), pp. 463-82. 

38. The London Polyglot was attacked by Dr. John Owen, the Puritan 
dean of Christ Church, Oxford, in Considerations on the Prolegomena and 
Appendix to the Late Polyglotta (1659), to which Walton made a prompt 
and sharp rejoinder, The Considerator Considered (1659). Walton's stature 
as a scholar and churchman was recognized at the Restoration, when he 
was appointed bishop of Chester in 1660. In 1667, the London Polyglot 
had the distinction of being put on the Index Librorum Prohibitorum. 

39- This Dr. Fell is the theme of Thomas Brown's well-known 
quatrain (adapted from Martial, Epigram i.32) beginning "I do not love 
thee, Dr. Fell." 

154 Tun Text of the New Testament 

Oxford. The text, drawn from the Elzevir 1633 edition, was supplied 
with an apparatus in which Fell claimed to give variants from more 
than 100 manuscripts and ancient versions. Unfortunately, however, 
about 20 of these witnesses, including Codex Vaticanus (B), are not 
cited individually but only in statements concerning the total number 
of manuscripts that agree in any particular reading. For the first 
time, evidence from the Gothic and Bohairic versions, supplied by 
T. Marshall, was also made available through Fell's apparatus. 

About the time of the publication of Fell's edition, John Mill 40 
(1645-1707), a fellow of Queen's College, Oxford, began his studies 
of New Testament textual criticism, which were to come to fruition 
30 years later in an epoch-making edition of the Greek text, pub- 
lished exactly 2 weeks before his death at the age of 62 (23 June 
1707). 41 Besides collecting all the evidence from Greek manuscripts, 
early versions, and fathers that lay within his power to procure, Mill 
prefixed to his edition valuable prolegomena in which he dealt with 
the canon of the New Testament and the transmission of the text, de- 
scribed 32 printed editions of the Greek Testament and nearly 100 
manuscripts, and discussed patristic citations from all the fathers of 
any importance. Some idea of the extent and detail of the prole- 
gomena may be had from the size of the index to the verses to which 
Mill makes reference in his discussion; these number 3,041 out of 
almost 8,000 verses in the whole New Testament. Despite the vast 
amount of solid learning embodied in his edition, however, Mill did 
not venture to form a text of his own but reprinted Stephanus' text 
of 1550 without intentional variation. 

In 1710, a reprint of Mill, with the prolegomena somewhat 
rearranged and with collations of 12 more manuscripts, was pub- 
lished at Amsterdam and Rotterdam by a Westphalian, I.udolf Kuster. 
Krister's reprint also appeared, with a new title page, at Leipzig in 
1723 and again at Amsterdam in 1746. 

As Walton's critical efforts had been attacked by Owen, so also 
Mill's monumental work came under fire from the controversial 
writer Dr. Daniel Whitby, rector of St. Edmund's, Salisbury. Alarmed by 
the great number of variant readings that Mill had collected — some 

40. Or Mills, as the Oxford Dictionary of the Christian Church 
prefers, s.v. 

41 . See Adam Fox, John Mill and Richard Bentley, a Study of the 
Textual Criticism of the New Testament, 1675-1729 (Oxford, 1954). 

The Precritical Period 155 

30,000 in all — Whitby argued that the authority of the holy Scriptures 
was in peril and that the assembling of critical evidence was tanta- 
mount to tampering with the text. 42 

There were others, however, who, appreciating the textual 
evidence collected by Mill, attempted to embody in practical form 
the results of his honest critical endeavors. Between 1709 and 1719, 
Dr. Edward Wells (1667-1727), a mathematician and theological 
writer, published at Oxford, in ten parts, a Greek Testament with a 
variety of helps for the reader. 43 Wells deserted the Elzevir text 
210 times, almost always agreeing with the judgment of nineteenth- 
century critical editors. Though Wells' edition was largely ignored by 
his contemporaries, history accords him the honor of being the first 
to edit a complete New Testament that abandoned the Textus 
Receptus in favor of readings from the more ancient manuscripts. 

42. Daniel Whitby, lixamen variantium lectionumj. Millii (London, 
1709). The English Deist Anthony Collins (1676-1729), did, in fact, appeal 
to the existence of so many variant readings as an argument against the 
authority of the Scriptures (A Discourse of Freethinking [London, 17131). 
The extent to which such considerations might be pushed is disclosed in 
Dean Swift's satirical essay "An Argument against the Abolition of Chris- 
tianity," in which he refers to a roue "who had heard of a text brought for 
proof of the Trinity, which in an ancient manuscript was differently read; 
he thereupon immediately took the hint, and by a sudden deduction of a 
long sorites, most logically concluded: 'Why, if it is as you say, I may safely 
whore and drink on, and defy the parson' " (Jonathan Swift, Works, iii 
[Edinburgh, 1814], p. 199). 

43- The content of Wells' edition may be seen from the lengthy 
descriptions on the title pages, of which the following is a sample: 

An Help for the more Easy and Clear Understanding of the Holy Scriptures: 
being the Two Sacred Treatises of St Luke, Viz. his Gospel and the Acts of the 
Apostles, Explained after the following Method, viz. 

I. The Original or Greek Text amended, according to the Best and most 

Ancient Readings. 
II. The Common English Translation rendered more Agreeable to the Original. 

III. A Paraphrase, wherein not only the Difficult Expressions and Passages 
are explain'd, but also Each Treatise is divided into Proper Sections and 
Paragraphs: and withall it is observ'd, What Supplements to the Two 
Gospels of St Matthew and Mark are given us by St Luke in his Gospel. To 
the End of each Treatise is subjoin'd a Synopsis of the Contents therof. 

IV. Annotations relating (as Occasion requires) to the Several Particulars. 
Oxford, 1719. 

156 The Text of the New Testament 

The name of Richard Bentley (1662-1742), master of Trinity Col- 
lege, Cambridge, is famous in the annals of classical scholarship for 
his exposure of the spurious Epistles of Phalaris, for his critical edi- 
tions of Horace and Terence, for his discovery of the use of the 
digamma in the Homeric poems, and generally for his skill in textual 
emendation. At an early age, Bentley began to correspond with var- 
ious scholars on the subject of a critical edition of the Greek and 
Latin New Testament. In 1720, he issued a six-page prospectus of 
printing such an edition, Proposals for Printing, giving as a specimen 
of his proposed text the last chapter of Revelation in Greek and 
Latin. 44 Here, Bentley abandoned the Textus Receptus in more than 
40 places. 45 

By following the oldest manuscripts of the Greek original and of 
Jerome's Vulgate, Bentley was confident that he could restore the 
text of the New Testament as it stood in the fourth century. 

By taking two thousand errors out of the Pope's Vulgate [Bentley refers 
to Pope Clement's edition of 1592], and as many out of the Protestant 
Pope Stephen's [referring to Stephanas' Greek text of 1550], I can set 
out an edition of each in columns, without using any book under nine 
hundred years old, that shall so exactly agree, word for word, and 
order for order, that no two tallies, nor two indentures, can agree 
better. 46 

It is obvious that the master of Trinity College was not inclined to 
underestimate his own abilities. In his Proposals, he refers to the 
forthcoming edition as "a XTfjfia eoaet, a charter, a Magna Charta, 
to the whole Christian church; to last when all the ancient MSS. here 
quoted may be lost and extinguished." 

In order to finance the publication, subscriptions were solicited 
and about £2,000 collected from considerably more than 1,000 
prospective purchasers of the edition. Despite the elaborate plans, 
however, and the amassing of new evidence from manuscripts and 

44. For a reproduction of the pamphlet, see Caspar Rene Gregory, 
Prolegomena (being vol. iii of Tischendorf 's Novum Testamentum Graece, 
ed. critica octava maior; Leipzig, 1884-94), pp. 231-40. 

45. Bentley's proposals were the occasion of an acrimonious contro- 
versy between Dr. Conyers Middleton and himself; for a summary of the 
points at issue, see James H. Monk, The Life of Richard Bentley, D.D., 2nd 
ed„ ii (London, 1833), pp. 130 ff. 

46. R. Bentley, Bentleii Critica Sacra, ed. by A. A. Ellis (Cambridge, 
1862), p. xv. 

The Precritical Period 157 

fathers, the scheme came to naught; and after Bentley's death, his 
literary executor returned the money to the subscribers. 47 

While Bentley was gathering materials for a definitive edition 
that would supplant the Textus Receptus, a Greek and English diglot 
in two volumes was published anonymously at London in 1729 with 
the title The New Testament in Greek and English. Containing the 
Original Text Corrected from the Authority of the Most Authentic 
Manuscripts: and a New Version Form 'd agreeably to the Illustrations 
of the most Learned Commentators and Critics-, with Notes and 
Various Readings, and a Copious Alphabetical Index. The edition has 
several typographical peculiarities. In the Greek text, the editor 
discards smooth breathing marks and accents; and at the close of 
questions, he uses a question mark instead of the Greek mark of 
interrogation (the semicolon). In the English translation, as well as in 
the explanatory notes, he begins sentences with a capital letter only 
at the beginning of a new paragraph. 

The editor of this diglot edition was Daniel Mace, a Presbyterian 
minister at Newbury, who chose from Mill's apparatus those variant 
readings that seemed to him to be superior to the Textus Receptus. 
In a high proportion of these alterations, Mace anticipated the opin- 
ions of much later scholars. 48 Likewise, his English translation reveals 
a certain independent vigor, for Mace adopted many racy and collo- 
quial expressions; for example, "don't," "can't," "what's," and (words 
of Simon the Pharisee to Jesus, Luke 7.40) "master, said he, lets hear 
it." Here and there he anticipated modern versions; for example, in 
Matt. 6.27, instead of the King James reference to adding one cubit 
to one's stature, Mace renders "who by all his sollicitude can add one 
moment to his age?" (In Luke 12.25, he translates "but which of you, 
with all his disquietude, can add one moment to the period of his 
life?") In footnotes and appended notes, Mace gives reasons for 
his departure from the earlier traditional text and translation. Several 
of these notes indicate Mace's free and independent spirit. Thus, in 
his note on Siva oooq eoxlv ev xn Agafiiq (Gal. 4.25), he declares: 
"This has all the marks of an interpolation: it is quite foreign to the 
argument, and serves only to perplex the apostle's reasoning, which 
without it appears very clear and coherent." He dismisses Mill's argu- 
ment in the "Prolegoma" § 1306 of his edition, Novum Testamentum 

47. See Fox, op. cit., pp. 105-26. 

48. For a list of some of these, see Eduard Reuss, Bibliotheca Novi 
Testamenti graeci . . . (Brunsvigae, 1872), pp. 175 f. 

158 The Tkxt of the New Testament 

Graecum, cum lectionibus variantibus (Oxford, 1707), based on the 
unanimity of manuscript evidence in its favor, with the contemptu- 
ous remark "as if there was any manuscript so old as Common Sense" 
(p. 689) and prints the conjecture to yag Ayag avoxoixei xn vvv 
Iegovoalnju. ... In his extended discussion of the authorship and 
the title of the Epistle to the Hebrews, he writes: 

A very learned writer of our own thinks xiii. 23 a sufficient proof that 
Paul was the original author, as if no body could be acquainted with 
Timothy but Paul, which shows, that in order to understand the doc- 
trine of Moral Evidence, that is, the doctrine of Chances, some other 
discipline and diet is necessary besides that of bearly chewing a few 
Hebrew roots, (p. 840) 

Like the work of many other innovators, Mace's edition was either 
vehemently attacked or quietly ignored. In England, Dr. Leonard 
Twells, vicar of St. Mary's in Marlborough, issued in three pans A Criti- 
cal Examination of the late New Testament and Version of the New 
Testament: wherein the Editor's Corrupt Text, False Version, and falla- 
cious Notes are Detected and Censur'd (London, 1731-2); 49 and on the 
Continent, scholars like Pritius, Baumgarten, and Masch rivaled Twells 
in their invective and abuse of Mace. But most theologians assumed an 
ostrich-like pose, and Mace's work was soon all but forgotten." 10 

With Johann Albrecht Bengel (1687-1752) we reach a new stage 
in the history of the textual criticism of the New Testament. 51 While 
a student in theology at Tubingen, his pietistic faith in the plenary 
inspiration of the Bible was disturbed by the 30,000 variants that had 
recently been published in Mill's edition of the Greek Testament, and 
he resolved to devote himself to the study of the transmission of the 
text. With characteristic energy and perseverance, he procured all 
the editions, manuscripts, and early translations available to him. 
After extended study, he came to the conclusions that the variant 
readings were fewer in number than might have been expected and 
that they did not shake any article of evangelic doctrine. 

49. In Part I, under the heading "False Renderings, and other foul 
Management favouring Arianism" (pp. 134-44), Twells lists 15 examples of 
what he, in some cases with justice, regarded as biased translation. 

50. See H. McLachlan, "An Almost Forgotten Pioneer in New 
Testament Criticism," Hibbert Journal, xxxvii (1938-9), pp. 617-25. 

51. The standard biography is Gottfried Malzer, Johann Albrecht 'Bengel, 
Leben unci Werk (Stuttgart, 1970); a recent, brief sketch is Lothar Bertsch, 
Johann Albrecht Bengel: Seine Lebensgeschicble (Hote.gerWngen, 2002). 

The Precritical Period 159 

In 1725, while teaching at the Lutheran preparatory school for 
ministerial candidates at Denkendorf, Bengel published an elaborate 
essay as a "forerunner" to his projected edition of the New Testament. 52 
Here, he laid down sound critical principles. He recognized that the 
witnesses to the text must not be counted but weighed, that is, clas- 
sified in "companies, families, tribes, nations." He was accordingly the 
first to distinguish two great groups, or "nations," of manuscripts: 
the Asiatic, which originated from Constantinople and its environs and 
included the manuscripts of more recent date, and the African, which 
he subdivided into two tribes, represented by Codex Alexandrinus 
and the Old Latin. For the weighing of variant readings, Bengel for- 
mulated a canon of criticism that, in one form or other, has been 
approved by all textual critics since. It is based on the recognition that 
a scribe is more likely to make a difficult construction easier than to 
make more difficult what was already easy. Formulated in Bengel's 
pithy Latin, it is proclivi scriptioni praestat ardua ("the difficult is to 
be preferred to the easy reading"). 

In 1734, Bengel published at Tubingen an edition of the Greek 
New Testament in a handsome quarto volume. He did not venture to 
correct the traditional Textus Receptus in accordance with personal 
judgment but followed (except in 19 passages in the Book of Reve- 
lation) the self-imposed rule of not printing any reading that had not 
been previously published in an earlier printed edition. He indicated 
in the margin, however, his views of the relative value of the variant 
readings according to the following categories: a designates the orig- 
inal reading; /J, a reading that is better than that which is printed in 
the text; y, a reading that is just as good as that in the text; 6, a reading 
that is less good than the text; e, a very inferior reading to be rejected. 
Bengel also took great pains to standardize the punctuation of the New 
Testament and to divide it into paragraphs, features that later editors 
borrowed from his edition. More than half of the volume is devoted 
to three excursuses in which Bengel supplied a reasoned account of 
his principles of textual criticism and an apparatus drawn from Mill's 
collations plus his own collations of 12 additional manuscripts. S3 

52. This essay is entitled "Prodromus Novi Testament! recte cauteque 
ordinandi" and was included as an appendix to his edition of Chrysostomi 
libri VI de sacerdotio (Denkendorf, 1725). 

53. A summary of his text-critical principles is available also in the 
preface to his celebrated commentary on the New Testament entitled 
Gnomon of the New Testament, translated by C. T. Lewis and M. R. Vincent, 
i (Philadelphia, I860), § viii. 

160 The Text of the New Testament 

Though Bengel was a man whose personal piety and life of 
good works were known to all (he had been in charge of an orphan 
home at Halle) and whose orthodoxy of belief was acknowledged 
(he was superintendent of the Evangelical Church of Wiirttemberg), 
he was treated as though he were an enemy of the holy Scriptures. 
So many persons impugned his motives and condemned his edition 
that he published in German, and then in Latin, a Defence of the New 
Testament. After Bengel's death, his son-in-law, Philip David Burk, 
published in 1763 an enlarged edition of the apparatus criticus, 
along with several short pamphlets Bengel had written to explain 
and defend his mature views on the correct methods of recovering 
the earliest form of the text of the New Testament. 

Among those who had collated manuscripts for Bentley was 
Johann Jakob Wettstein (1693-1754), a native of Basle, where he be- 
came a Protestant minister. His taste for textual criticism showed it- 
self early; when ordained to the ministry at the age of 20, he deliv- 
ered an address on variant readings in the New Testament. His 
textual studies, however, were interpreted by some as preparations 
for denying the doctrine of the divinity of Christ; and in 1730, he was 
deposed from his pastorate and driven into exile. From 1733 on- 
ward, he was professor of philosophy and Hebrew in the Arminian 
college at Amsterdam (in succession to the celebrated Jean Leclerc), 
where he resumed his textual studies. 54 In 1751-2, the fruits of 
40 years of research appeared in his publication at Amsterdam of a 
magnificent edition of the Greek New Testament in two folio vol- 
umes. Though he printed the Elzevir text, he indicated in the margin 
those readings that he himself held to be correct. In an appendix en- 
titled "Animadversiones et cautiones ad examen Variarum Lectionum 
N. T. necessariae,"' 55 Wettstein sets forth a good deal of sound 
advice; for example, he states that codices autem pondere, non 
numero estimandi sunt (§ xviii fin., "manuscripts must be evaluated 
by their weight, not by their number"). Despite his generally excel- 
lent theoretical views, Wettstein was somewhat haphazard in apply- 
ing his rules. Furthermore, he came to advocate (largely, it seems, in 

54. On Wettstein's wide-ranging lectures under the general rubric of 
"philosophy," see C. L. Hulbert-Powell's biography, John James Wettstein, 
1693-1754 (London, 1938), pp. 196 f. 

55. Vol. ii, pp. 851-74, reproduced, wiuh some condensation, from his 
anonymously published treatise Prolegomena adNovi Testamenti graeci 
editionem accuratissimam . . . (Amsterdam, 1730), pp. 165-201. 

The Precritical Period 161 

opposition to Bengel) the quite untenable theory that all of the early 
Greek manuscripts have been contaminated by the Latin versions 
and that, consequently, the later Greek manuscripts should be relied 
upon as preserving a more authentic text. 

Wettstein's apparatus is the first in which the majuscule manu- 
scripts were regularly denoted by capital Roman letters and the mi- 
nuscule manuscripts (including lectionaries), by Arabic numerals — a 
system that has continued to be used to the present time. In addition 
to the textual material, Wettstein's edition provides a thesaurus of 
quotations from Greek, Latin, and rabbinical authors, illustrating the 
usage of words and phrases of the New Testament. Though his 
critical judgment was not as sound as Bengel's, his passion for the 
study of manuscripts, which took him on extensive journeys, re- 
sulted in the collation or recollation of about 100 manuscripts, and 
his commentary is still a valuable storehouse of classical, patristic, 
and rabbinical lore. 5 

Though he published no edition of the Greek Testament, Johann 
Salomo Semler (1725-91), often regarded as the father of German ra- 
tionalism, made noteworthy contributions to the science of textual 
criticism by his reprint of Wettstein's Prolegomena, with discerning 
comments of his own. 57 Adopting Bengel's system of classifying 
manuscripts by groups, Semler carried the process still further by as- 
signing the origin of Bengel's Asiatic group (which he renamed 
"Eastern") to the recension 58 prepared in the early fourth century by 
Lucian of Antioch, and the origin of Bengel's African group (Semler's 
"Western" or "Egypto-Palestinian" group) to Origen. Subsequently, 
Semler expanded his textual researches and classified New Testa- 
ment manuscripts in three recensions: (1) Alexandrian, derived from 
Origen and preserved in the Syriac, Bohairic, and Ethiopic versions; 
(2) Eastern, current in the Antiochian and Constantinopolitan 

56. In an appendix to his edition of the Greek Testament, Wettstein 
published the editio princeps of the Syriac text of the two pseudo- 
Clementine epistles, De virginitate, with a Latin translation in parallel 

57. J. S. Semler, Wetstenii libelli ad crisin atque interpretationem Novi 
Testamenti (Halle, 1764). 

58. Semler was the first to apply the term recension to groups of New 
Testament witnesses {Hermeneutiscbe Vorbereitung, iii 11] [Halle, 17651). 
Properly, a recension is the result of deliberate critical work by an editor; 

it is, however, often used in a loose sense as synonymous with family. 

162 The Text of the New Testament 

Churches; and (3) Western, embodied in the Latin versions and 
fathers. Later witnesses, he thought, 'were characterized by a mixture 
of all recensions. 59 

For the next important edition of the Greek Testament, we must 
return to England, where William Bowyer, Jr. (1699-1777), was the 
third generation in a line of famous printers in London. Often 
regarded as "the most learned English printer of whom we have any 
account," 60 Bowyer not only exercised a scholarly vigilance in print- 
ing a wide variety of volumes but frequently contributed learned pref- 
aces, annotations, and corrigenda to the works that passed through 
his publishing house. After his father and he had issued several 
editions of the Textus Receptus of the Greek Testament (in 1715, 
1728, 1743, and 1760), Bowyer decided to produce a critical edition 
worthy of the reputation of his firm. In 1763, such an edition was 
issued in two volumes of duodecimo size. Bowyer constructed his 
text largely by following the critical judgments that Wettstein had 
expressed in his marginal notes regarding the earliest form of text. 
Using square brackets, Bowyer marked in his text not a few familiar 
passages that lack the support of good manuscripts; for example, he 
bracketed the doxology of the Lord's Prayer (Matt. 6.13), thepericope 
deadultera (John 7.53-8.11), the Comma Johanneum (1 John 5.7-8), 
and single verses (such as Acts 8.37 and 15.34) and words through- 
out the New Testament. In other passages, Bowyer departed from 
the Textus Receptus by introducing into his edition readings that the 
better manuscripts support. 61 In an appendix to the second volume, 
Bowyer included nearly 200 pages of conjectural emendations on 
the New Testament, bearing on the text and punctuation. 62 

59. Idem, Apparatus ad liberalem Novi Testamenti interpretationem 
(Halle, 1767). 

60. S. Austin Allibone, A Critical Dictionary of English Literature, i 
(Philadelphia, 1871), p. 229. 

61. For a selected list of these along with other details of Rowyer's 
edition, reference may be made to B. M. Metzger, Chapters in the History 
of New Testament Textual Criticism (Grand Rapids, MI, 1963), pp. 155-60. 

62. A second edition of the Conjectures, with extensive additions, was 
published separately at London in 1772 with the following title, Critical 
Conjectures and Observations on the New Testament, Collected from Various 
Authors, as Well in Regard to Words as Pointing: With the Reasons on Which 
Both are Founded. This volume was translated into German by Joh. Chr. F. 
Schulz (2 vols., Leipzig, 1774—5). Two other editions, enlarged still further, 
were published posthumously at London, one in 1782 and the other in 
1812, the latter from an annotated, interleaved copy of Dr. Henry Owen. 

The Precritical Period 163 

Another Englishman, Edward Harwood (1729-94), a Noncon- 
formist minister, published at London in 1776 a two-volume edition 
of the New Testament that was, according to the statement on the 
title page: 

Collated with the most approved Manuscripts; with Select Notes in 
English, critical and explanatory; and References to those Authors who 
have best illustrated the Sacred Writings. To which are added, a 
Catalogue of the principal Editions of the Greek Testament; and a List 
of the most esteemed Commentators and critics. 

For the Gospels and Acts, Harwood followed in the main the 
text of Codex Bezae and for the Pauline letters, Codex Claromon- 
tanus. Where these were not available, he utilized other manuscripts, 
chiefly Codex Alexandrinus. 63 In an analysis of 1,000 passages in the 
New Testament, Reuss found that Harwood deserted the Textus 
Receptus more than 70% of the time and, in 643 passages, agreed 
with the epoch-making critical edition of Lachmann, published in the 
nineteenth century (see pp. 170-1). 

It is not surprising that, in a period when the Textus Receptus 
held sway and only occasionally an independent spirit ventured to 
question its authority, the first Greek Testament to be published in 
America was the time-honored Textus Receptus. 64 The printer of this 
edition, Isaiah Thomas, Jr. (1749-1831), was a typically enterprising 

63. In his preface, Harwood declares: 

Excepting typographical errors, which a moderate acquaintance with the lan- 
guage will easily enable the reader to correct, I persuade myself, that the Text 
of the inspired writers here exhibited will approve itself to every Scholar who 
is a judge of sacred criticism, to be as near to the original autograph of the 
Evangelists and Apostles as any hitherto published in the world. To accom- 
plish this arduous design, I carefully read through the late Professor Wetstein's 
Greek Testament, published at Amsterdam in two Volumes in folio, scrupu- 
lously weighed the merit or demerit of the various lections there exhibited 
from a great multitude of Manuscripts of different value, and adopted only 
those which to my judgment appeared to be best authenticated: my meaning 
is, that I espoused only those which I verily believed to be the very words 
which the inspired authors originally wrote, (pp. viii-ix) 

64. The first Bible to be printed in America was the translation made 
by John Eliot into the Algonquin language, published at Cambridge, 
Massachusetts, in 1661-3. The first Bible in a European language to be 
printed in America was Luther's German translation published in 1743 by 
Christoph Sauer in Germantown, Pennsylvania. The first Bible in English to 
be published in America came from the press of Robert Aitken at 
Philadelphia in 1782. 

164 The Text of the New Testament 

and hard-working Yankee. Apprenticed to a printer at the age of 6, 
after only 6 weeks of indifferent schooling, Thomas began his up- 
ward climb that resulted in his becoming a member of nearly every 
learned society in the country and the recipient of honorary degrees 
from Dartmouth and Allegheny Colleges. His printing establishment 
issued over 900 different books, more than those of Benjamin 
Franklin, Hugh Gaine, and Matthew Carey, his nearest rivals. Indeed, 
owing to the excellence of his typographical work and the range and 
number of his imprints, Franklin called him "the Baskerville of 
America." Sensing that the market called for an edition of the Greek 
New Testament, Thomas secured the assistance of a scholarly minister, 
Rev. Caleb Alexander, 65 and issued the editio prima Americana of the 
Greek New Testament. The volume, a small duodecimo of 478 pages, 
was published at Worcester, Massachusetts, in April 1800. 

The title page states that the edition reproduces accurately the 
text of John Mill's edition (juxta exemplar Joannis Millii accuratissime 
impressum). This is, however, not entirely true to fact, for in more than 
a score of passages the editorial work of Alexander can be detected. 
According to Isaac H. Hall, 66 a comparison with editions issued by 
Beza and by the Elzevirs shows that Alexander, in eclectic fashion, 
occasionally chose a reading now from this edition and now from that. 
Externally, the format of the volume bears many resemblances to 
Bowyer's editions of the Textus Receptus. In fact, the title page of the 
Alexander-Thomas edition reproduces exactly, line for line, word for 
word, and style for style of type (except only the date and name and 
place of publisher), the title page of Bowyer's edition of 1794. 67 

65. Caleb Alexander (1755-1828), a native of Northfield, Massachu- 
setts, was graduated from Yale College in 1777. During his pastorate at 
Mendon, a village not far from Worcester, Alexander found time to write 
two Latin grammars and a Greek grammar, the latter being published by 
Thomas with the title Grammatical System of the Grecian Language 
(Worcester, MA, 1796). It was doubtless at this period that Thomas 
secured Alexander's services in supervising the preparation of a Greek 
New Testament. 

66. Isaac H. Hall, A Critical Bibliography of the Greek New Testament 
as Published in America (Philadelphia, 1883), p. 11. 

67. For further information regarding the publisher and the text of this 
first American edition of the Greek New Testament, see B. M. Metzger, 
"Three Learned Printers and Their Unsung Contributions to Biblical Schol- 
arship," Journal of Religion, xxxii (1952), pp. 254-62. 


The Modern Critical Period 

From Griesbach to the Present 

I. The Beginnings of Scientific Textual 
Criticism of the New Testament 

During the latter part of the eighteenth century, the German scholar 
Johann Jakob Griesbach (1745-1812) laid foundations for all subse- 
quent work on the Greek text of the New Testament. A pupil of 
Semler's at Halle, Griesbach was professor of the New Testament at 
the University of Jena from 1775 until his death. After traveling in 
England, Holland, and France in order to collate manuscripts, he 
devoted special attention to the New Testament quotations in the 
Greek fathers and to several versions of the New Testament that pre- 
viously had been little studied, such as the Gothic, the Armenian, 
and the Philoxenian Syriac. 

Griesbach also investigated the history of the transmission of the 
New Testament text in antiquity and further developed Bengel's and 
Semler's grouping of manuscripts in recensions. At first, he was in- 
clined to divide the extant materials into five or six different groups; 
he afterward limited them to three.- the Alexandrian, Western, and 
Byzantine recensions. The standard of the Alexandrian text he be- 
lieved to be Origen, who, though writing many of his works in 
Palestine, was assumed to have brought with him into exile copies 
of the Scriptures similar to those used in his native city. To this 


166 The Text of the New Testament 

group Griesbach assigned the majuscule manuscripts C, L, and K; 
the minuscules 1, 13, 33, 69, 106, and 118; the Bohairic, Armenian, 
Ethiopic, and Harclean Syriac; and, in addition to quotations of 
Origen, those of Clement of Alexandria, Eusebius, Cyril of Alexandria, 
and Isidore of Pelusium. To the Western group he assigned Codex D, 
the Latin versions, and, in pan, the Peshitta Syriac and Arabic versions. 
The Constantinopolitan group, which he regarded as a later compi- 
lation from the other two, was represented by A (in the Gospels) and 
by the great mass of later majuscule and minuscule manuscripts as 
well as the larger proportion of patristic quotations. 

Among the 15 canons of textual criticism that Griesbach elabo- 
rated, the following (his first canon) may be given as a specimen: 

The shorter reading (unless it lacks entirely the authority of the ancient 
and weighty witnesses) is to be preferred to the more verbose, for 
scribes were much more prone to add than to omit. They scarcely ever 
deliberately omitted anything, but they added many things; certainly 
they omitted some things by accident, but likewise not a few things 
have been added to the text by scribes through errors of the eye, ear, 
memory, imagination, and judgement. Particularly the shorter reading 
is to be preferred, even though according to the authority of the wit- 
nesses it may appear to be inferior to the other, — 

a. if at the same time it is more difficult, more obscure, ambiguous, 
elliptical, hebraizing, or solecistic; 

b. if the same thing is expressed with different phrases in various man- 

c. if the order of words varies; 

d. if at the beginning of pericopes; 

e. if the longer reading savours of a gloss or interpretation, or agrees 
with the wording of parallel passages, or seems to have come from 

But on the other hand the longer reading is to be preferred to the 
shorter (unless the latter appears in many good witnesses), — 

a. if the occasion of the omission can be attributed to homoeoteleuton; 

b. if that which was omitted could have seemed to the scribe to be 
obscure, harsh, superfluous, unusual, paradoxical, offensive to pious 
ears, erroneous, or in opposition to parallel passages; 

c. if that which is lacking could be lacking without harming the 
sense or the structure of the sentence, as for example incidental, 
brief propositions, and other matter the absence of which would 
be scarcely noticed by the scribe when re-reading what he had 

The Modern Critical Period 167 

d. if the shorter reading is less in accord with the character, style, or 
scope of the author; 

e. if the shorter reading utterly lacks sense; 

f. if it is probable that the shorter reading has crept in from parallel 
passages or from lectionaries. 

Griesbach showed great skill and tact in evaluating the evidence 
of variant readings. For example, his judgment, based on patristic and 
versional evidence, that the shorter form of the Lord's Prayer in Luke 
11.3-4 is to be preferred was remarkably confirmed a few years later 
when the readings of Codex Vaticanus were published, for it was 
found that all of the omissions are supported by that early manuscript. 

The importance of Griesbach for New Testament textual criti- 
cism can scarcely be overestimated. For the first time in Germany a 
scholar ventured to abandon the Textus Receptus at many places 
and to print the text of the New Testament in the form to which his 
investigations had brought him. Though at times Griesbach permit- 
ted himself to be led astray by a too mechanical adherence to his sys- 
tem of recensions, 1 his textual labors on the whole were character- 
ized by caution and candor. His principal editions were published at 
Halle in 1775-7, at Halle and London in 1796-1806, and at Leipzig in 
1803-7. Several editions of his text were also issued by enterprising 
printers in England, Scotland, and America. His influence was ex- 
tended still further when his work was adopted as the basis of 
smaller manual editions issued on the Continent by Schott, Knapp, 
Tittmann, Hahn, and Theile. 

Soon after the publication of Griesbach 's first edition, several 
other scholars published collations that greatly increased the avail- 
ability of evidence for the text of the New Testament from Greek 
manuscripts, the early versions, and the Church fathers. Christian 
Friedrich Matthaei (1744-1811), professor first at Wittenberg and 
then at Moscow, where he taught classical literature, issued at Riga 
in 12 parts, between 1782 and 1788, an edition of the Greek text with 
the Latin Vulgate. His printed text is of little value because it is based 
on manuscripts of recent date, but his apparatus is valuable. Besides 

1 . He was vehemently criticized for this and other reasons by 
Archbishop R. Laurence in his Remarks on the Systematical Classification 
of Manuscripts Adopted by Griesbach in His Edition of the New Testament 
(Oxford, 1814; repr. in Biblical Repertory, ed. by Charles Hodge, ii [18261, 
pp. 33-95). 

168 The Text of the New Testament 

collating manuscripts at Dresden, Leipzig, and Gottingen, Matthaei 
sought out biblical and patristic manuscripts in Moscow, originally 
brought to Russia from Mount Athos. 2 He made, for example, colla- 
tions of 34 manuscripts of the homilies of John Chrysostom on the 
Gospels and the Pauline Epistles. In the second edition of his New 
Testament, without the Latin Vulgate (3 vols., 1803-7), Matthaei pro- 
vides evidence from about 30 additional manuscripts. His edition is 
noteworthy for containing, apparently for the first time, evidence 
from the Slavic version of the New Testament. In an appendix to his 
edition of the Book of Revelation, Matthaei lists ten Slavic manu- 
scripts that he had examined; he contented himself, however, with 
collating the text of Revelation in the folio edition of the Slavic Bible 
published at Moscow in 1762. This evidence is given in Latin, the 
collation having been made against a manuscript of the Latin Vulgate 
at Moscow, Codex Demidovianus. 

Franz Karl Alter (1749-1804), a Jesuit of Silesia who became pro- 
fessor of Greek at Vienna, published an edition of the Greek Testa- 
ment in two volumes (Vienna, 1786-7), based on the text of a single 
manuscript in the Imperial Library at Vienna. In separate appendices, 
he cited evidence from 22 other Greek manuscripts, two Latin man- 
uscripts, the Bohairic version (which David Wilkins had edited at 
Oxford in 1716), and four Slavic manuscripts. This is the first edition 
of the Greek New Testament that contained evidence from Slavic 
manuscripts themselves. 

A still larger addition to the quantity of materials available to the 
textual critic was made through the efforts of four Danish scholars, 
Andreas Birch, Jacob G. C. Adler, D. G. Moldenhauer, and O. G. 
Tychsen, who were sent by the king of Denmark, Christian VII, to 
examine manuscripts in Italy, Spain, Germany, and other European 
countries. The results of their labors were published by Birch in a vol- 
ume describing the then known Greek manuscripts of the New Testa- 
ment 3 and in four volumes of collations (Copenhagen, 1788-1801). 

2. While in Russia, Matthaei managed to steal a good many manu- 
scripts of both the classics and the fathers. Some of these he kept in his 
own library, while others he sold or gave to various libraries and friends in 
Germany and Holland. For an account of his life with incriminating evi- 
dence of his brazen thievery, see Oscar von Gebhardt in Centralblatt fur 
Bibliothekswesen, xv (1898), pp. 345-57, 393-420, 441-82, and 537-66. 

3. Andreas Birch, Kritisk Beskrivelse over groeske Haandskrifter af det 
Nye Testamente (Copenhagen, 1785). 

The Modern Critical Period 169 

The latter contain variant readings from 172 Greek manuscripts and 
evidence from two Syriac versions (the Philoxenian and Palestinian). 
Many of the manuscripts, however, were only partially examined by 
Birch and his colleagues, including Codex Vaticanus (B), readings of 
which now for the first time appeared in print. 

About this time, two Roman Catholic scholars gave, in different 
ways, an impetus to the textual criticism of the New Testament. 
Johann Leonhard Hug (1765-1846), professor at the University of 
Freiburg im Breisgau, developed the theory 4 that at about the begin- 
ning of the third century the several types of New Testament text 
degenerated rapidly and produced what is commonly called the 
Western text, which Hug called the xoivt) ixdomg (common edition). 
Toward the middle of the third century, according to Hug, this 
edition was revised in Palestine by Origen, a revision adopted later 
by Jerome; in Egypt, it was revised by Hesychius and in Syria, by 
Lucian, a presbyter of Antioch, both of which revisions Jerome con- 
demned. Although Hug started from what was on the whole a true 
conception of the Western text and its manifold variations, his inge- 
nious attempt to connect three recensions of the Septuagint (whose 
places of origin he believed assured) with three types of New Testa- 
ment text failed. Johannes Martin Augustinus Scholz (1794-1852), a 
pupil of Hug's and professor at the University of Bonn, traveled 
extensively throughout Europe and the Near East in order to draw 
up what was the first comprehensive listing of Greek manuscripts of 
the New Testament, adding 6l6 new manuscripts to those previously 
known. He was the first to emphasize the importance of ascertaining 
the geographical provenance represented by the several manuscripts, 
a point that B. H. Streeter was to elaborate in 1924 by his theory of 
"local texts." Unlike Streeter, who relied on the congruence of manu- 
script readings with patristic citations, Scholz was guided chiefly by 
certain external signs of provenance, such as details of paleography, 
iconography, marginal notes, colophons, and evidence regarding 
local saints who were honored in lectionaries. 

After some tentative attempts at classifying manuscripts, Scholz 
came to adopt essentially Bengel's division into two families, which 
he called the Alexandrian and the Constantinopolitan. During his 
extensive examinations of minuscule manuscripts, he was impressed 

4. J. L. Hug, FAnleitung in die Scbriften des Neuen Testaments (Stuttgart, 
1808), §§ 22 ff. 

170 The Text of the New Testament 

by their general uniformity of text type, a feature which he regarded 
as evidence of their superiority to the earlier Alexandrian type. Thus, 
Scholz's two- volume edition of the Greek Testament (Leipzig, 1830-6) 
marked a retrogression in textual criticism toward the Textus Recep- 
tus; only here and there does it happen to contain readings supported 
by the earlier manuscripts because the editor was inconsistent in the 
application of his critical theories. It is symptomatic of the low ebb to 
which appreciation of textual scholarship had sunk in England at this 
time that Scholz's edition was welcomed and praised by many British 
scholars and its text reprinted by Bagster in London in several edi- 
tions. At a later date (1845), Scholz retracted his preference for the 
Byzantine text and declared that if a new edition of his Greek Testa- 
ment were called for, he would receive into the text most of the 
Alexandrian readings that he had formerly placed in the margin. 

II. The Overthrow of the Textus Receptus 

The first recognized scholar to break totally with the Textus 
Receptus was the celebrated classical and Germanic philologist of 
Berlin Karl Lachmann (1793-1851), who published an edition of the 
Greek Testament that rests wholly upon the application of textual 
criticism in the evaluation of variant readings. Lachmann is famous 
for his editions of ancient classical authors, including Propertius, 
Catullus, Tibullus, Lucretius, as well as medieval epics and lyrics 
such as the Nibelungenlied, Walther von der Vogelweide, and 
Wolfram von Eschenbach. He demonstrated how, by comparison of 
manuscripts, it is possible to draw inferences as to their lost ances- 
tors or archetypes, their condition, and their pagination. In his most 
famous work, that on Lucretius, he showed that the peculiarities of 
the three chief manuscripts all derive from a single archetype, con- 
taining 302 pages of 26 lines each, and thus he was able to make 
various transpositions in the received text. 

In editing the New Testament, Lachmann's aim was not to re- 
produce the original text, which he believed to be an impossible 
task, but to present on purely documentary evidence, apart from any 
previously printed editions, the text current in Eastern Christendom 
at the end of the fourth century (about A.D. 380). Using no minus- 
cule manuscripts, he based his text on several of the earlier majus- 
cules, the Old Latin and Jerome's Vulgate, and the testimony of 
Irenaeus, Origen, Cyprian, Hilary, and Lucifer. After 5 years of work, 

The Modern Critical Period 171 

in 1831 he published at Berlin an edition of the Greek text, with a list 
of passages where it differs from the Textus Receptus. Brackets are 
used to indicate words of doubtful textual authority. Instead of in- 
cluding in the edition itself an account of his methodology and the 
reasons that led him to reject the Textus Receptus, Lachmann chose 
to refer the reader to an article that he had published the previous 
year in a German periodical. 5 It is not surprising that theologians, 
even liberal ones like de Wette, generally misunderstood Lachmann's 
intentions and attacked him with considerable vehemence, coining 
such names of reproach as "Bentley's ape" (simia Bentleit). In the 
preface to his second edition (2 vols., Berlin, 1842-50), Lachmann 
replied in kind, arrogantly twitting his critics for their blind prefer- 
ence for the familiar but corrupt later text to the earlier, purer form. 
It was not always appreciated that Lachmann did not pretend to 
print the original text of the New Testament but only a provisional 
one, namely, that current in the fourth century, including even pal- 
pable scribal errors if sufficiently well attested. The weakness of the 
edition is the slender manuscript basis to which Lachmann restricted 
himself. According to Scrivener, "Lachmann's text seldom rests on 
more than four Greek codices, very often on three, not infrequently 
on two; in Matt. 6.20-8.5, and in 165 out of 405 verses of the 
Apocalypse, on but one." 6 Despite such limitations, however, most 
later scholars have agreed with Westcott and Hort's evaluation of 
Lachmann and his work: 

A new period began in 1831, when for the first time a text was con- 
structed directly from the ancient documents without the intervention 
of any printed edition, and when the first systematic attempt was made 
to substitute scientific method for arbitrary choice in the discrimination 
of various readings. In both respects the editor, Lachmann, rejoiced to 
declare that he was carrying out the principles and unfulfilled inten- 
tions of Bentley, as set forth in 1716 and 1720. 7 

5. "Rechenschaft iiber seine Ausgabe des Neuen Testaments," Theolo- 
gische Studien unci Kritiken, iii (1830), pp. 817^5. 

6. F. H. A. Scrivener, A Plain Introduction to the Criticism of the New 
Testament, 4th ed., ii (London, 1894), p. 233. 

7. B. F. Westcott and F. J. A. Hort, The New Testament in the Original 
Greek, [ii,] Introduction [and] Appendix (London, 1881), p. 13- For a 
twentieth-century assessment of Lachmann and his work, see Sebastiano 
Timpanaro, La genesi del metodo del Lachmann (Florence, 1963). 

172 The Text of the New Testament 

The man to whom modern textual critics of the New Testament 
owe most is without doubt Lobegott Friedrich Constantin von 
Teschendorf (1815-74), who sought out and published more manu- 
scripts and produced more critical editions of the Greek Bible than 
any other single scholar. Between 1841 and 1872 he prepared eight 
editions of the Greek Testament, some of which were reissued alone 
or with German or Latin versions, as well as 22 volumes of texts of 
biblical manuscripts. The total number of his books and articles, 
most of them relating to biblical criticism, exceeds 150. 8 

While studying theology at Leipzig from 1834 to 1838, young 
Tischendorf came under the influence of Johann G. B. Winer, whose 
grammar of New Testament Greek (1822) went through many edi- 
tions and remained the standard for several generations. Winer in- 
fused in his student a passion to seek and to utilize the most ancient 
witnesses in reconstructing the purest form of the Greek Scriptures. 
To this task the young scholar dedicated himself; writing to his 
fiancee, he declared: "I am confronted with a sacred task, the strug- 
gle to regain the original form of the New Testament." At the age 
of 25, supported by a small stipend from the government of Saxony, 
Tischendorf began the laborious work of deciphering the palimpsest 
Codex Ephraemi and certain other manuscripts in the Bibliotheque 
Nationale at Paris. Subsequently, he visited libraries throughout 
Europe and the Near East, searching for and examining manuscripts 
new and old. (For the story of his discovery of Codex Sinaiticus, see 
pp. 62-4.) 

Of Tischendorf s several editions of the Greek Testament, the 
most important is the eighth (editio octava critica maior), issued in 
11 parts, beginning in 1864, and published in two volumes 
(Leipzig, 1869-72). This was accompanied by a rich apparatus 
criticus in which he assembled all of the variant readings that he or 
his predecessors had found in manuscripts, versions, and fathers. 
Soon after the publication of the second volume, a stroke of palsy 
prevented Tischendorf from continuing his labors. A third volume 
of valuable prolegomena to the edition was prepared by Caspar 

8. See the biographical article on Tischendorf by Caspar Rene 
Gregory in Bibliotheca Sacra, xxxiii (1876), pp. 153-93, containing a list of 
Tischendorf s publications, as well as the assessment of Matthew Black 
and Robert Davidson, Constantin von Tischendorf and the Greek New 
Testament (Glasgow, 1981). 

The Modern Critical Period 173 

Rene Gregory and issued in three parts (Leipzig, 1884, 1890, 
1894). 9 

Tischendorfs claim to fame rests chiefly upon his indefatigable 
industry in assembling textual evidence; his use of that evidence 
in constructing his editions, however, was marked by a somewhat 
wooden adherence to a number of critical canons as well as a certain 
arbitrariness in dealing with problems not covered by the canons. 
The text of his eighth edition differs (according to Eberhard Nestle) 
from the seventh edition in 3,572 places, and he has been accused of 
giving excessive weight to the evidence of Codex Sinaiticus, which 
he had discovered between issuing the two editions. 

In England, the scholar who, at the middle of the nineteenth cen- 
tury, was most successful in drawing British preference away from 
the Textus Receptus was Samuel Prideaux Tregelles (1813-75). As a 
boy, he had shown exceptional talent and intellectual curiosity and, 
while earning his livelihood at an ironworks, managed to devote his 
spare time to learning Greek, Aramaic, Hebrew, and Welsh. While 
still in his early twenties, Tregelles began to form plans for a new crit- 
ical edition of the New Testament. Having observed how persistently 
Scholz rejected the evidence of the earliest manuscripts and being 
dissatisfied with the somewhat hesitating way in which Griesbach 
still clung to the Textus Receptus, he determined to employ his 
leisure time in preparing an edition based only on the evidence of 
the earliest witnesses. Without his knowing it, Tregelles developed 
critical principles that paralleled to a remarkable degree those of 
Lachmann. Thereafter, he was engaged in the collation of Greek 
manuscripts, making extensive travels throughout Europe for the 
purpose. His careful and systematic examination of practically all 
the then known majuscules and several of the important minuscules 
resulted in the correction of many erroneous citations made by pre- 
vious editors. He also examined afresh the New Testament quota- 
tions found in Greek Church fathers down to Eusebius as well as 
the ancient versions and edited (186l) a palimpsest manuscript of the 
Gospel of Luke, Codex Zacynthius (£"), acquired in 1821 by the 
British and Foreign Bible Society. Before issuing any portion of his 
new text, however, Tregelles published a survey of earlier editions, 

9. Gregory's volume of prolegomena, with additions and corrections, 
was later published in German in three parts, entitled Textkritik des Neuen 
Testamentes (Leipzig, 1900-9). 

174 The Text of thk New Testament 

in which he set forth his own critical principles (An Account of the 
Printed Text of the Greek New Testament . . . , London, 1854); he also 
rewrote that portion of T. H. Home's encyclopedic Introduction to 
the Critical Study and Knowledge of the Holy Scriptures, which relates 
to the textual criticism of the New Testament (vol. iv, 10th ed.; London, 

Unlike Tischendorf, who hurried into print with another edi- 
tion as soon as he had discovered some new manuscript evidence, 
Tregelles preferred to fix his full energy upon the final goal of a de- 
finitive text representing his mature judgment and issued but one 
edition. This was published at London in six parts between 1857 and 
1872. Disabled by a stroke of paralysis in 1870, he secured the assis- 
tance of B. W. Newton for the final fascicle. A volume of prolegom- 
ena compiled from Tregelles' other works and containing many 
pages of addenda et corrigenda was edited by F. J. A. Hort and A. W. 
Streane and published posthumously in 1879. In spite of poverty, 
opposition, and ill health, Tregelles overcame all difficulties and de- 
voted a lifetime of meticulous labors to the text of the New Testa- 
ment as an act of worship, undertaken, as he declares in the preface, 
"in the full belief that it would be for the service of God, by serving 
His Church." 

Though remembered primarily for his widely used commentary 
on the New Testament, Henry Alford (1810-71), dean of Canterbury 
and author of several well-known hymns (among them "Come, Ye 
Thankful People, Come" and "Ten Thousand Times Ten Thousand"), 
deserves mention here as an ardent advocate of the critical principles 
formulated by those who, like Lachmann, had worked for the 
"demolition of the unworthy and pedantic reverence for the received 
text, which stood in the way of all chance of discovering the genuine 
word of God." 10 In the successive editions of his commentary, Alford 
set forth more and more fully the evidence of variant readings and 
boldly printed that form of Greek text that he believed was sup- 
ported by the earliest and best witnesses. 

The year 1881 was marked by the publication of the most note- 
worthy critical edition of the Greek Testament ever produced by 
British scholarship. After working about 28 years on this edition 
(from about 1853 to 1881), Brooke Foss Westcott (1825-1901), canon 

10. Henry Alford, The Greek Testament, with a Critically Revised 
Text . . . , new ed., i (New York, 1881), p. 76 of the prolegomena. 

The Modern Critical Period 175 

of Peterborough and regius professor of divinity at Cambridge (con- 
secrated bishop of Durham in 1890), and Fenton John Anthony Hort 
(1828-92), Hulsean Professor of Divinity at Cambridge, issued two 
volumes entitled The New Testament in the Original Greek. Volume i 
contains the Greek text; volume ii comprises a valuable introduction 
and appendix, in which the critical principles followed by the two 
editors are set forth in detail by Hort, 11 with the concurrence of his 
colleague, and certain problem passages are discussed. Occasion- 
ally, when the editors could not agree on certain details, the opinion 
of each is identified by his initials. The second edition of the second 
volume, published in 1896, contains some additional notes by F. C. 
Burkitt on the recently discovered Sinaitic Syriac manuscript. 12 

Unlike earlier editors, neither Westcott nor Hort was concerned 
to collate manuscripts, nor did they provide a critical apparatus. 
Rather, utilizing previous collections of variant readings, they refined 
the critical methodology developed by Griesbach, Lachmann, and 
others and applied it rigorously, but with discrimination, to the wit- 
nesses to the text of the New Testament. The principles and proce- 
dures of criticism that they elaborated may be summarized as follows. 

Hort begins the classic "Introduction" by discussing what he calls 
"Internal Evidence of Readings." 

The most rudimentary form of criticism consists in dealing with each 
variation independently, and adopting at once in each case out of two 
or more variants that which looks most probable. . . . Internal Evidence 
of Readings is of two kinds, which cannot be too sharply distinguished 
from each other; appealing respectively to Intrinsic Probability, having 
reference to the author, and what may be called Transcriptional Proba- 
bility, having reference to the copyists. In appealing to the first, we ask 
what an author is likely to have written: in appealing to the second, we 
ask what copyists are likely to have made him seem to write. 13 

11. For further information about Hon, see E. G. Rupp, Hort and the 
Cambridge Tradition (Cambridge, 1970); Graham A. Patrick, F.J. A. Hort, 
Eminent Victorian (Sheffield, 1988); Graham Neville, "Science and Tradi- 
tion: F.J. A. Hort and His Critics," Journal of Theological Studies, n.s., 1 
(1999), pp. 560-82. 

12. In 1988, Hendrickson Publishers (Peabody, MA) issued a reprint 
of vol. ii, "Introduction" and "Appendix," but unfortunately did not repro- 
duce a copy of the second edition [18961 with Burkitt's "Notes on Select 

13. Westcott and Hort, op. cit., pp. 19-20. 

176 The Text of the New Testament 

When, as sometimes happens, Intrinsic and Transcriptional Probabil- 
ities are in conflict, it is usually safer to make judgments on the basis 
of what Hort called the "observed proclivities of average copyists" than 
on what one imagines the original author must have written. 

In order to transcend the limitations inherent in a procedure 
based solely on Internal Evidence of Readings, the textual critic must 
also utilize Internal Evidence of Documents. When weighing the 
evidence in individual cases, one gains assurance by considering 
whether a witness is normally credible and trustworthy. Therefore, 
instead of being content with evaluating one reading after another, 
in isolation from each other, the critic should collect information 
regarding the character of individual manuscripts. If one finds that a 
given manuscript frequently supports certain readings that clearly 
commend themselves as original on the basis of probability, it is nat- 
ural to prefer its readings in other instances when the Internal Evi- 
dence of Readings is not clear enough for a decision. Hort summarizes 
this point by enunciating the principle that "knowledge of documents 
should precede final judgement upon readings." 14 

The next step involves examination of the relationship of the 
several witnesses to one another. Manuscripts may be grouped and 
considered from the standpoint of their genealogy. If, for example, of 
ten manuscripts, nine agree against one but the nine have a common 
original, the numerical preponderance counts for nothing. The clear- 
est evidence in tracing the genealogy of witnesses is the presence of 
conflate readings, that is, readings that have arisen from the combi- 
nation of elements that had existed previously in separate manu- 
scripts. Here, Hort enunciates another principle of criticism, that "all 
trustworthy restoration of corrupted texts is founded on the study of 
their history, that is, of the relations of descent or affinity which con- 
nect the several documents." 15 

Finally, in his discussion of methodology, Hort considers the 
Internal Evidence of Groups, which is in some sense intermediate 
between the Internal Evidence of Documents and the Genealogical 
Evidence. Just as it is useful to determine the general characteristics 
of a given manuscript by observing how often it supports or rejects 
readings that have been previously evaluated individually on the 
basis of Internal Probability, so the general characteristics of a given 

14. Ibid., p. 31. 

15. Ibid., p. 40. 

The Modern Critical Period 177 

group of witnesses can be determined and evaluated in relation to 
other groups. 

The validity of inferences based on this procedure depends on 
the genealogical principle that "community of reading implies com- 
munity of origin." 16 Such generalizations on the value of groups of 
witnesses, in turn, assist the critic in coming to decisions when mix- 
ture in the ancestry of manuscripts makes it difficult to draw up a 

The paragraphs above contain a summary of the critical princi- 
ples adopted and elaborated by Westcott and Hort. The results of 
their application of these principles to the then known New Testa- 
ment manuscripts will now be briefly set forth. 

On the basis of investigations into the relationships among the 
witnesses to the text of the New Testament, Westcott and Hort distin- 
guished four principal types of text: the Syrian, Western, Alexandrian, 
and Neutral. 

1. The latest of these four forms of text is the Syrian, which is a 
mixed text resulting from a revision made by an editor or editors in 
the fourth century who wished to produce a smooth, easy, and com- 
plete text. This conflated text, the farthest removed from the originals, 
was taken to Constantinople, whence it was disseminated widely 
throughout the Byzantine Empire. It is best represented today by 
Codex Alexandrinus (in the Gospels, not in Acts and the Epistles), the 
later majuscule manuscripts, and the great mass of minuscule manu- 
scripts. The Textus Receptus is the latest form of the Syrian text. 

Hort's classic description of the Syrian text is as follows: 

The qualities that the authors of the Syrian text seem to have most 
desired to impress on it are lucidity and completeness. They were 
evidently anxious to remove all stumbling-blocks out of the way of the 
ordinary reader, so far as this could be done without recourse to vio- 
lent measures. They were apparently equally desirous that he should 
have the benefit of instructive matter contained in all the existing texts, 
provided it did not confuse the context or introduce seeming contra- 
dictions. New omissions accordingly are rare, and where they occur are 
usually found to contribute to apparent simplicity. New interpolations 
on the other hand are abundant, most of them being due to harmonistic 
or other assimilation, fortunately capricious and incomplete. Both in 
matter and in diction the Syrian text is conspicuously a full text. It 

16. Ibid., p. 60. 

178 The Text of toe New Testament 

delights in pronouns, conjunctions, and expletives and supplied links of 
all kinds, as well as in more considerable additions. As distinguished 
from the bold vigour of the "Western" scribes, and the refined scholar- 
ship of the Alexandrians, the spirit of its own corrections is at once sen- 
sible and feeble. Entirely blameless on either literary or religious 
grounds as regards vulgarised or unworthy diction, yet shewing no 
marks of either critical or spiritual insight, it presents the New Testa- 
ment in a form smooth and attractive, but appreciably impoverished in 
sense and force, more fitted for cursory perusal or recitation than for 
repeated and diligent study. 17 

2. Of the remaining types of text that Westcott and Hort isolated, 
the so-called Western type is both ancient and widespread. It is pre- 
served in certain bilingual majuscule manuscripts, notably Codex 
Bezae of the Gospels and Acts (D) and Codex Claromontanus of the 
Epistles (D p ), the Old Latin version(s), and the Curetonian Syriac. 
Its date of origin must have been extremely early, perhaps before 
the middle of the second century. Marcion, Tatian, Justin, Irenaeus, 
Hippolytus, Tertullian, and Cyprian all made use to a greater or 
lesser extent of a Western form of text. 

One of the marked characteristics of the Western text, according 
to Hort, is a love of paraphrase: 

Words, clauses, and even whole sentences were changed, omitted, and 
inserted with astonishing freedom, wherever it seemed that the mean- 
ing could be brought out with greater force and definiteness. . . . An- 
other equally important characteristic is a disposition to enrich the text 
at the cost of its purity by alterations or additions taken from traditional 
and perhaps from apocryphal or other non-biblical sources. [The West- 
ern text is also characterized by] the multiplication of genitive pronouns, 
but occasionally their suppression where they appeared cumbrous; the 
insertion of objects, genitive, dative, or accusative, after verbs used ab- 
solutely, the insertion of conjunctions in sentences that had none, but 
occasionally their excision where their force was not perceived and the 
form of the sentence or context seemed to commend abruptness; free 
interchange of conjunctions; free interchange of the formulae introduc- 
tory to spoken words; free interchange of participle and finite verb with 
two finite verbs connected by a conjunction; substitution of compound 
verbs for simple as a rule, but conversely where the compound verb of 
the true text was difficult or unusual; and substitution of aorists for 
imperfects as a rule, but with a few examples of the converse. . . . 

17. Ibid., pp. 134 f. 

The Modern Critical Period 179 

Another impulse of scribes abundantly exemplified in Western 
readings is the fondness for assimilation. In its most obvious form it is 
merely local, abolishing diversities of diction where the same subject 
matter recurs as part of two or more neighbouring clauses or verses, or 
correcting apparent defects of symmetry. But its most dangerous work 
is "harmonistic" corruption, that is, the partial or total obliteration of 
differences in passages otherwise more or less resembling each other. 18 

3. The Alexandrian text, according to Westcott and Hort, is pre- 
served to a greater or lesser extent in Codex Ephraemi (C), Codex 
Regius (L), Codex 33, and the Coptic versions (especially the 
Bohairic), as well as the quotations of the Alexandrian fathers 
Clement, Origen, Dionysius, Didymus, and Cyril. Its characteristic is 
that which might be expected from the influence of a Greek literary 
center, a delicate philological tact in correcting forms and syntax and 
in subtle changes made in the interest of attaining a greater degree 
of polish in language and style (such as the rearrangement of the 
order of words to avoid hiatus). 

4. The Neutral text, as its question-begging name implies, is, in 
the opinion of Westcott and Hort, the most free from later corruption 
and mixture and the nearest to the text of the autographs. It is best 
represented by Codex Vaticanus (B) and next by Codex Sinaiticus 00. 
The concurrence of these two manuscripts is very strong and shows 
that they cannot be far from the original text. With the exception of a 
few passages, which they specify, Westcott and Hort declare: 

It is our belief (1) that the readings of 8 B should be accepted as the 
true readings until strong internal evidence is found to the contrary, 
and (2) that no readings of t« B can safely be rejected absolutely, 
though it is sometimes right to place them only on an alternative foot- 
ing, especially where they receive no support from Versions or 
Fathers. 19 

The exceptions to their preference for the Neutral text are several 
passages that they term "Western non-interpolations." They doubtless 
chose this cumbersome nomenclature simply because they could not 
bring themselves to refer directly to "Neutral interpolations," which 
is exactly what, on their own reconstruction, is involved in these 
readings. In several passages in the last three chapters of Luke and 

18. Ibid., pp. 122-4. 
19- Ibid., p. 225. 

180 The Text of the New Testament 

one in Matthew, 20 the Western text is regarded by Westcott and Hort 
as preserving the original form. The reason they abandon the testi- 
mony of N and B in these passages is that here the Western text, 
which normally is the fuller and more circumstantial form, has re- 
sisted (so they believe) the impulse to add material, whereas it is the 
Neutral text that presents the expanded reading. 

In accordance with Westcott and Hort's critical reconstruction, 
the relation of their four text types to the autograph may be repre- 
sented by the following stemma: 


WESTERN ^. ' (Common ancestor of both 

(D Old Lat Syr 1 - ^5^ I Alexandrian and Neutral 

fam. so far as 


SYRIAN ((A) E F G H S V Y most minuscules) 


By way of retrospect and evaluation, it may be said that scholars 
today generally agree that one of the chief contributions made by 
Westcott and Hort was their clear demonstration that the Syrian (or 
Byzantine) text is later than the other types of text. Three main types 
of evidence support this judgment: (1) the Syrian text contains com- 
bined or conflate readings that are clearly composed of elements 
current in earlier forms of text; (2) no ante-Nicene father quotes a 

20. The Western noninterpolations, which Westcott and Hort print 
within double brackets, are in Matt. 27.49 and Luke 22.19-20, 24.3, 6, 12, 
36, 40, 51, and 52. Nearly a score of other, somewhat similar passages 
throughout the Gospels form an intermediate class that, in their opinion, 
may also involve Western noninterpolations; see ibid., p. 176. 

The Modern Critical Period 181 

distinctively Syrian reading; and (3) when the Syrian readings are 
compared with the rival readings, their claim to be regarded as orig- 
inal is found gradually to diminish and at last to disappear. 21 

It was perhaps not surprising that Westcott and Hort's total rejec- 
tion of the claims of the Textus Receptus to be the original text of the 
New Testament should have been viewed with alarm by many in the 
church. During the closing decades of the nineteenth century, the 
traditional text found a doughty defender in the person of John W. 
Burgon (1813-88), dean of Chichester. He has been described as 
"a High-churchman of the old school," who became notorious as 
"a leading champion of lost causes and impossible beliefs; but the 
vehemence of his advocacy somewhat impaired its effect." 22 His 
conservatism can be gauged from a sermon he preached at Oxford in 
1884 in which he denounced the higher education of "young women 
as young men" as "a thing inexpedient and immodest"; the occasion 
was the admission of women to university examinations! 

The publication in 1881 of the Revised Version of the King 
James, or Authorized, Version of 1611 aroused Burgon's indignation 
not only on the score of its unidiomatic English but even more be- 
cause the revisers had followed an underlying Greek text substan- 
tially identical with that of Westcott and Hort. In a series of three 
learned articles in the (London) Quarterly Review, which were 
reprinted in a volume entitled The Revision Revised (London, 1883), 
Burgon used every rhetorical device at his disposal to attack both the 
English revision and the Greek Testament of Westcott and Hort. 
Burgon's argument was basically theological and speculative. As an 
ardent high churchman, he could not imagine that, if the words of 
Scripture had been dictated by the inspiration of the Holy Spirit, God 
would not have providentially prevented them from being seriously 
corrupted during the course of their transmission. Consequently, it 
was inconceivable to Burgon that the Textus Receptus, which had 
been used by the Church for centuries, could be in need of the dras- 
tic revision that Westcott and Hort had administered to it. 

What Burgon was apparently unable to comprehend was the 
force of the genealogical method, by which the later, conflated text is 
demonstrated to be secondary and corrupt. Instead of following the 

21. Ibid., pp. 93-119. 

22. So the Oxford Dictionary of National Biography, ed. by H. C. G. 
Matthew and Brian Harrison, 8 (Oxford, 2004), p. 805. 

1 82 The Text of the New Testament 

text of the few earlier manuscripts, Burgon preferred the readings 
supported by the majority of the later witnesses/ 3 Consequently, so 
far from sharing Westcott and Hort's high regard for the testimony of 
Codex Vaticanus and Codex Sinaiticus, Burgon maintained that, with 
the single exception of D, which exhibits the wildest text of all, the two 
manuscripts honored by Westcott and Hort are the most depraved. 
He assures his readers 

without a particle of hesitation, that nBD are three of the most scan- 
dalously corrupt copies extant: — exhibit the most shamefully mutilated 
texts which are anywhere to be met with: — have become, by whatever 
process (for their history is wholly unknown), the depositories of the 
largest amount of fabricated readings, ancient blunders, and inten- 
tional perversions of Truth, — which are discoverable in any known 
copies of the Word of GoD. 2i 

23- Burgon found an ally in Thomas Ft. Birks, honorary canon of Ely, 
in his Essay on the Right Estimation of Manuscript Evidence in the Text of 
the New Testament (London, 1878). Birks, who attempted to assign a math- 
ematical weight to individual manuscripts, considered the later manuscripts 
in general to be more valuable than the earlier ones! 

The anachronistic views of Burgon were resuscitated by Edward F. 
Hills in his booklet Ibe King fames Version Defended! A Christian View of 
the New Testament Manuscripts (Des Moines, 1A, 1956), in which the au- 
thor outdid Burgon in defending the Textus Receptus, arguing even for the 
genuineness of the Comma Johanneum of 1 John 5.7-8. See also Hills' 
introduction "Dean Burgon in the Light of Criticism" in the 1959 reprint of 
Burgon's book The Last Twelve Verses of the Gospel According to St. Mark 
Vindicated. . . (Oxford, 1871). 

24. The Revision Revised (London, 1883), p. 16 (italics are Burgon's). 

At the suggestion of Prebendary Edward Miller, Burgon's literary 
executor, a debate was held at New College, Oxford, on 6 May 1897, 
when the position of Burgon was upheld by Miller, G. H. Gwilliam, and 
A. Bonus against William Sanday, A. C. Headlam, and Willoughby C. Allen, 
who advocated the textual views of Westcott and Hort. One of the chief 
points of contention was the date of the Peshitta Syriac version of the New 
Testament. Miller maintained that this version, which is a witness to the 
Syrian type of text, goes back to the second century and that, therefore, 
the Syrian type of text did not originate with Lucian and his contempo- 
raries at the beginning of the fourth century. Sanday acknowledged that 
the date of the Peshitta was "the sheet anchor" of Miller's position but 
was unable to produce convincing evidence for its later origin ( The 
Oxford Debate on the Textual Criticism of the New Testament [London, 
18971, p. 28). 

The Modern Critical Period 183 

Two other British scholars, F. H, A. Scrivener and George Salmon, 
were also critical of Westcott and Hort's theories but were far more 
temperate than Burgon in the expression of their dissent. The former 
objected to Hort's total rejection of the testimony of the Syrian text, 25 
and the latter complained that more consideration should have been 
given to the weight of purely Western readings. 2 

This somewhat lengthy account of the work of Westcott and 
Hort may be concluded with the observation that the overwhelming 
consensus of scholarly opinion recognizes that their critical edition 
was truly epoch-making. They presented what was doubtless the 
oldest and purest text that could be attained on the basis of infor- 
mation available in their day. Though the discovery of additional 
manuscripts has required the realignment of certain groups of 
witnesses, 27 the general validity of their critical principles and proce- 
dures is widely acknowledged by textual scholars today. 

During his long and fruitful life, Bernhard Weiss (1827-1918), 
professor of New Testament exegesis at Kiel and at Berlin, edited the 
New Testament in Greek (3 vols., Leipzig, 1894-1900; 2nd, small ed., 
3 vols., 1902-5). Primarily an exegete, Weiss brought to his task an 
extensive and detailed knowledge of the theological and literary 
problems of the text of the New Testament. Instead of grouping 
manuscript authorities and evaluating variants in terms of external 
support, Weiss discriminated among readings in accordance with 
what he deemed to be the most appropriate meaning in the context. 
His procedure was to go through each New Testament book with a 
critical apparatus and to consider important textual variants, select- 
ing in each case that reading which seemed to him to be justified, as 
Hort would have said, by intrinsic probability. While this procedure 
is certainly subjective to an extreme, one must not suppose that other 
methods are entirely objective. Even Westcott and Hort's criticism is 

A few years later, F. C. Burkitt set forth evidence in his monograph 
St. Ephraim's Quotations from the Gospel (Cambridge, 1901), showing that 
the New Testament quotations in the genuine works of Ephraem (d. 373) 
agree largely with the Old Syriac version and not with the Peshitta and 
that therefore the latter appears to have been produced subsequently to 
Ephraem's lifetime. 

25. Scrivener, op. cit., pp. 287 f. 

26. George Salmon, Some Thoughts on the Textual Criticism of the 
New Testament (London, 1897), pp. 129 ff. 

27. See pp. 276 ff. 

184 The Text of thf New Testament 

subjective, for first they chose the method they decided to follow and 
then they judged, largely on the grounds of intrinsic and transcriptional 
probabilities, that their so-called Neutral text is generally to be pre- 
ferred above all other types of text. 

After Weiss had edited his text by adopting the variants he re- 
garded as most appropriate to the author's style and theology, he 
drew up lists of different classes of error that he observed among the 
variant readings and evaluated each of the chief Greek manuscripts 
in accordance with its relative freedom from such faults. The classes 
of error that he detected are: (1) harmonizations among the Gospels, 
(2) the interchange of words, (3) omissions and additions, (4) alter- 
ations of word order, and (5 ) orthographical variation. In assessing 
the degree of freedom of Greek manuscripts from these errors, Weiss 
decided that Codex Vaticanus was the best. It is not surprising, there- 
fore, that the general complexion of Weiss's edition is remarkably 
similar to that of Westcott and Ilort's, who relied so largely on Codex 
Vaticanus. The importance of Weiss's text is not only that it repre- 
sents the mature opinion of a great exegetical scholar who had given 
years of detailed consideration to the meaning of the text but also 
that the results of his subjective methodology confirm the results of 
scholars who followed a different procedure, sometimes regarded as 
more objective because it started from the grouping of the manu- 
scripts themselves. 28 

Though issued in 1910 (the text remained unchanged in the 
second edition, 1947), Alexander Souter's Greek New Testament 
represents the position of British textual scholarship as it was in 
1881; the edition merely reproduces the Greek text that Archdeacon 
Edwin Palmer, a member of the New Testament panel of British 
translators, constructed as the text that inferentially lies behind the 
Revised Version of 1881. Taking the third edition of Stephanus (1550) 
as the basis of his edition, Palmer formed a continuous text that 
represents the decisions of the revisers. When, however, the English 
revision was considered to represent correctly either of two com- 
peting readings, Palmer did not ordinarily alter the Textus Receptus. 
As a result, the orthography, the spelling of proper names, and the 

28. See C. R. Gregory, "Bernhard Weiss and the New Testament," 
American Journal of Theology, i (1896), pp. 16-37; Kirsopp Lake, 
"Dr. Weiss's Text of the Gospels: The Thoughts of a Textual Critic on the 
Text of an Exegete," American Journal of Theology, vii (1903), pp. 249-58. 

The Modern Critical Period 185 

typographical peculiarities or errors of Stephanus are, with a few 
exceptions, retained. 

Souter's contribution in 1910 was a selected critical apparatus to 
go with Palmer's text. The chief strength of this apparatus lies in the 
relatively full evidence that is quoted from the Church fathers, par- 
ticularly the Latin fathers. In 1947, the apparatus was enlarged by the 
addition of evidence from the Chester Beatty papyri and other wit- 
nesses brought to light since 1910. As regards textual complexion, 
Souter's edition is closer to the Textus Receptus than is any other 
widely used Greek Testament today. 29 

A noteworthy edition of the Greek New Testament that ap- 
peared in the first part of the twentieth century was von Soden's 
Die Schriften des Neuen Testaments in ihrer altesten erreichbaren 
Textgestalt hergestellt auf Grund ihrer Textgeschichte, I. Teil, Unter- 
suchungen (Berlin, 1902-10); II. Teil, Text mit Apparat (Gottingen, 
1913). 30 Hermann Freiherr von Soden was born in 1852 at Cincinnati, 
Ohio, and died in 1914 as the result of a mishap that occurred while 
boarding a subway train in Berlin. Containing the results of pro- 
longed investigation of the Greek minuscules and of intensive study 
of the history of the Greek text, von Soden's edition has nevertheless 
been described as "a magnificent failure." 

Through the financial assistance of a friend, Elise Koenigs, von 
Soden was able to send out a considerable number of research 
students and scholars to examine manuscripts in the libraries of 
Europe and the Near East. These helpers secured partial or complete 

29. On the basis of an analysis of 11 sample chapters scattered 
throughout the New Testament, J. Harold Greenlee found that Nestle 's text 
differs from the Textus Receptus in 233 cases, Merk's text in 160 cases, 
Bover's text in 111 cases, Vogels' text in 67 cases, and Souter's in 47 cases. 
In terms of percentage, this means that Nestle differs from the Textus 
Receptus 496% of the time in comparison with Souter! See Kurt Aland, 
who quotes Greenlee's statistics in "The Position of New Testament Textual 
Criticism" in Studia Evangelica, ed. by Aland, F. L. Cross, et al. (Texte und 
Untersuchungen, lxxiii, Berlin, 1959), p. 719. For more extensive statistics 
pertaining to the comparison of several modern editions of the Greek 
New Testament, see Kurt Aland, Studien zur Uberliefemng des Neuen 
Testaments und seines Textes (Berlin, 1967), pp. 59-80. 

30. von Soden also published a smaller edition (Handausgabe), enti- 
tled Griechisches Neues Testament, Text mit kurzem Apparat (Gottingen, 

186 The Text of the New Testament 

collations of an immense number of hitherto unexamined manu- 
scripts. Utilizing this information, von Soden set forth his views re- 
garding the history of the text in 2,203 pages of prolegomena, much 
of it printed in small type! Dissatisfied with the sigla previously used 
to designate majuscule and minuscule manuscripts, he devised a 
new system of nomenclature that indicates the age, content, and 
type of each manuscript. Though the system is ingenious, it is also 
so highly complicated 31 that most subsequent textual critics have 
refused to adopt it, preferring instead the old system, which 
was slightly revised by Gregory 32 in order to remove a number of 
anomalies. As a result, in order to make practical use of von Soden's 

31. Von Soden divides all known Greek manuscripts into three 
classes: (1) the d manuscripts, containing the whole New Testament 
(.dia.dijxrj'), with or without the exception of the Book of Revelation; 

(2) the £ manuscripts, containing the Gospels (evayyekiov); (3) the a manu- 
scripts, containing Acts and the Epistles, with or without the Book of Revela- 
tion (cbrocTtoAoc) . Within each of these classes, numbers are assigned in 
accordance with the date and contents of each manuscript. The 6 and a 
manuscripts, up to the close of the ninth century, are numbered 1-49; 
those of the tenth century, 50-90; for the following centuries, numbers of 
three digits are used and the numeral in the hundreds place indicates the 
century (thus, 146 is a manuscript of the eleventh century; 446, a manu- 
script of the fourteenth century). In 6 manuscripts, the presence of the 
Book of Revelation is indicated by using 1 — 49 in each hundred and 50-99 
for those without it (thus, d 421 would be a fourteenth-century manuscript 
containing the whole of the New Testament; d 271 would be a twelfth- 
century manuscript containing all the books of the New Testament except 
the Apocalypse). Similarly, the contents of the a manuscripts are indicated 
by a still more involved system of numerals. Since the fi manuscripts are 
so very numerous, the system is modified by the addition of other digits. 
Besides being intolerably complicated, von Soden's nomenclature tells 
more than one is likely to want to know — for what useful end is gained by 
being informed of the other New Testament books that a manuscript con- 
tains in addition to the book in which the given variant appears? Further- 
more, the system permits the addition of only a limited number of newly 
found manuscripts, and makes no provision at all for the Greek lectionar- 
ies. Moreover, since opinion concerning the date of individual manuscripts 
may change, the elaborate provision to represent the age of a given docu- 
ment may well perpetuate what has come to be regarded as an error. 

32. Caspar Rene Gregory, Die griechischen Handscbriften des Neuen 
Testaments (Leipzig, 1908). 

The Modern Critical Period 187 

apparatus, it is necessary to consult a "key" to unlock the signifi- 
cance of what otherwise are meaningless hieroglyphs. 33 

Von Soden's classification of the text types of manuscripts of the 
Gospels is based on considerations of their general textual character- 
istics, on the form of the text of the pericope de adultera, and on the 
chapter divisions attached to them. Using these criteria, he divided 
the witnesses into three main groups: the Koine, the Hesychian, and 
the Jerusalem recensions. 

The KiKoivtf) text is divided into about 17 subgroups, of which 
K 1 is the oldest and best form. Produced by Lucian of Antioch (mar- 
tyred A.D. 312), this, with various subsequent alterations, became 
the prevailing text throughout the Byzantine Church (Westcott and 
Hort's Syrian text). 

The H CHov%iog) text, which von Soden traced to Hesychius of 
Egypt, is preserved today in the old majuscules (B, «, C, L, A, and W), 
some minuscules (33, 579, 892, and 1241), the Sahidic and Bohairic 
versions, and the Alexandrian fathers Athanasius, Didymus, Cyril, and 
others. It therefore includes what Westcott and Hort called the Neutral 
and Alexandrian texts. 

The / Clegooo^vfia) text, deriving probably from Eusebius and 
Pamphilus of Caesarea in Palestine, is not preserved in substantial 
integrity in any outstanding manuscripts but must be elicited from a 
number of authorities of mixed characteristics. The best witnesses 
in the Gospels are the majuscules D and and the minuscules 28, 
372, 565, and 700; but so diverse are the textual phenomena that 
von Soden was compelled to posit 17 sub-groups of witnesses that 
are more or less closely related to this text. 

According to von Soden, these three recensions go back to a 
lost archetype, the I-H-K text, used by Origen but already 
corrupted in the second century by Marcion, in the case of the 
Pauline Epistles, and by Tatian, in the case of the Gospels and Acts. 
The discovery and elimination of these corruptions bring us to the 
original text. 

33. For convenient keys to transpose von Soden's system of enumera- 
tion into the Gregory system, see Friedrich Kriiger, Schliissel zu von Sodens 
Die Schriften des Neuen Testaments (Gottingen, 1927); Benedikt Kraft, 
Die Zeichenfur die wichtigeren Handschriften des griechischen Neuen 
Testaments, 3te Aufl. (Freiburg, i. Br., 1955). 

188 The Text of the New Testament 

Among the principles that von Soden followed in constructing 
his text are the following: 

1. When the readings of the three great recensions are certain, the 
reading supported by two recensions is generally adopted. 

2. If two recensions have a reading that agrees with a parallel, 
the reading of the third that differs from the parallel is usually 

3. The reading supported by Tatian is at once open to the suspicion 
of departing from the original text. Only in the event of two re- 
censions agreeing with Tatian and the dissenting recension agree- 
ing with a parallel is the latter to be adjudged secondary, and this 
remains the case even when the former reading also agrees with 
a parallel. 

4. When early, certainly mutually independent witnesses; — even 
though they may be only patristic writers or versions — agree in a 
reading that differs from Tatian, this reading requires serious con- 
sideration for adoption even when all three recensions agree with 

While acknowledging the enormous amount of research that 
von Soden's edition represents, most scholars have criticized his 
methods and results in the following respects: 34 

1. Since von Soden tends to give preference to readings supported 
by two of the three main texts, by this procedure the Koine type 
of text is elevated to a rank coordinate in importance with the 
other two texts. So far from regarding the Koine as an indepen- 
dent entity, however, most scholars today follow the view of 
Griesbach, Hort, and others, that this text is largely secondary and 
derivative from the others. As a consequence of von Soden's high 
estimate of the value of the Koine text, his edition approaches the 
Textus Receptus more closely than does any previous modern 
critical text. 35 

34. See, e.g., K. Lake in Review of Theology and Philosophy, iv (1908), 
pp. 201-17, 277-95; H. C. Hoskier \n Journal of Theological Studies, xv 
(1914), pp. 307-26; Hans Lietzmann in Zeitschrift fur die neutestamentliche 
Wissenschaft, xv (1914), pp. 323-31; A. Souter in Expositor, 8th sen, x 
(1915), pp. 429-44. 

35. Compare what is said concerning Souter's edition (see pp. 184-5), 
which does not pretend to be a critically established Greek text. 

The Modern Critical Period 189 

2. Though von Soden thought that his chief contribution to textual 
studies was the isolation and subdivision of his / text, later schol- 
ars regard it as the least sound, for he includes in one text type 
such heterogeneous elements as the Western witnesses, the Cae- 
sarean text, the Old Latin, and the Old Syriac, as well as witnesses 
that are mixed with the Koine text. 

3. While Marcion and Tatian undoubtedly had a certain corrupting 
influence upon the transmission of the New Testament text, von 
Soden assigned them an altogether disproportionate degree of 
importance in the contamination not only of the Latin and Syriac 
versions but of Greek witnesses as well. 

4. Though absolute accuracy in an extensive critical apparatus is 
probably unattainable, where von Soden's work can be tested, it 
has been found to contain a higher percentage of errors than is 
usually considered to be consistent with trustworthy scholarship. 

Despite these and other justifiable criticisms that have been lev- 
eled against von Soden, his edition remains a monument of broad 
research and immense industry that, with the extensive prolegomena 
dealing with the history of the transmission of the text, must be taken 
into account by every serious textual critic. 

The next three editions to be mentioned are the products of 
twentieth-century Roman Catholic scholarship. The edition prepared 
by Heinrich Joseph Vogels (Dusseldorf, 1920; with the Latin Vulgate, 
1922; 4th ed., Freiburg, 1955) is closer to the Textus Receptus than 
the other two. 36 The editor provides a limited apparatus which, in 
addition to citing the principal majuscule and minuscule manu- 
scripts, is relatively full as regards evidence derived from the Old 
Latin materials and the Syriac versions. 

Relying on von Soden's apparatus but transposing von Soden's 
sigla to those of the Gregory system and adding fresh manuscript 
evidence, Augustin Merk, S. J., published an edition of the Greek 
and Latin New Testament through the Pontifical Biblical Institute 
(Rome, 1933; 6th ed., 1948; 7th ed., 1951, S. Lyonnet; 8th ed., 1958, J. P. 
Smith; 9th ed., 1964, C. M. Martini; 10th ed. 1984). 37 The apparatus, 

36. See n. 29. 

37. Merk's text was reprinted by Gianfranco Nolli in his Novum 
Testamentum graece et latine (Rome, 1955). The footnotes of Nolli's edition 
supply the student with a limited number of text-critical, syntactical, and 
lexical aids. 

190 The Text of the New Testament 

which includes evidence from the several Tatianic witnesses, is drawn 
up so as to show family relationship among the witnesses. Unfortu- 
nately, however, Merk's citation of evidence is far from accurate;* 
and when his apparatus supplies evidence not available for verifi- 
cation in other publications, one hesitates to rely upon his testimony. 
In the construction of his Greek text, Merk departs further from the 
Textus Receptus than do the other two Roman Catholic editors. 

Jose Maria Bover, S.J., devoted his efforts over many years to the 
collection and evaluation of textual materials. 39 The Greek text of 
his bilingual edition (Madrid, 1943; 6th ed., 1981), which is printed 
with the beautiful font of Greek type belonging to the Association 
Guillaume Bude, is an eclectic one, departing frequently from the 
Alexandrian type of text and approaching the Western or Caesarean 
type. The apparatus, which presents information concerning the tex- 
tual opinions of six modern editors, supplies manuscript evidence 
for only the more important variants. 

A widely used pocket edition of the Greek Testament was pre- 
pared by Eberhard Nestle (1851-191 3)' i0 for the Wurttembergische 
Bibelanstalt (Stuttgart, 1898; 24th ed., I960, by Erwin Nestle 
[1883-19721 and Kurt Aland [1915-1994]). Its text (since the 3rd ed., 
1901) was based on a comparison of the texts edited by Tischendorf 
(1869-72), Westcott and I Iort (1881), and Bernhard Weiss (1894-1900); 
where two of these three editions agree, this reading was printed by 
Nestle.'* 1 Thus, the text of Nestle represented the state of nineteenth- 
century scholarship; its apparatus, however, which is a marvel of con- 
densation, supplies with a high degree of accuracy a great amount of 

38. See, e.g., H. G. Opitz's review in Gnomon, xii (1936), pp. 429-36; 
G. D. Kilpatrick, "Three Recent Editions of the Greek New Testament," 
Journal of Theological Studies, 1 (1949), pp. 145 ff. 

39. For a summary of Bover's textual investigations, see B. M. Metzger, 
Chapters in the History of New Testament Textual Criticism (Grand Rapids. 
MI, 1963), pp. 121—41. 

40. See Warren A. Kay, "The Life and Work of Eberhard Nestle," in 
The Bible as Book: The Transmission of the Greek Text, ed. by Scot 
McKendrick and Orlaith A. O'Sullivan (London, 2003), pp. 187-99. 

41. Since the 17th ed. of 1941, however, a small number of variants 
that, according to widespread scholarly opinion, have strong claims to be 
original have been taken into the text against the majority of the three 
nineteenth-century editions. Aland's name first appears on the title page of 
Nestle in the 21st ed., 1952. 

The Modern Critical Period 191 

textual information, including many early witnesses that were discov- 
ered during the twentieth century. Starting with the 26th edition of the 
Nestle-Aland Novum Testamentum Graece (1979), the text (though 
not the apparatus) is identical with that found in the United Bible 
Societies' Greek New Testament (see n. 47). 

In celebration of 100 years (1898-1998) since the publication of 
the Nestle Novum Testamentum Graece, the Deutsche Bibelge- 
sellschaft issued a jubilee edition, described as "einmalige, limitierte 
Jubilaums-Ausgabe." This is a copy of the 27th edition, with 15 pages 
of introduction (in German and English) by Barbara Aland and Beate 
Koster and an appendix of plates, two of Eberhard Nestle (c. 1875 
and c. 1912), one of Erwin Nestle in the 1920s, and one of Kurt Aland 
at the beginning of the 1980s. 

In connection with the sesquicentennial celebration of the 
British and Foreign Bible Society (1804-1954), a new edition of 
Eberhard Nestle's 1904 text was edited with an apparatus prepared 
by G. D. Kilpatrick with the help of Erwin Nestle and several other 
scholars (London, 1958). The text of the 1904 edition was changed 
in about 20 passages (of which 1 1 are listed in the introduction), and 
several alterations were made in orthography, accentuation, and the 
use of brackets. As regards the apparatus, the number of variants 
cited is substantially smaller than in current editions of Nestle, but a 
certain amount of additional information is provided for the variants 
that are cited. (For comments on Kilpatrick's Greek text of the Diglot 
New Testament, privately circulated by the British and Foreign Bible 
Society, see p. 225.) 

In addition, reference should be made to the apparatus criticus 
published by S. C. E. Legg at Oxford in 1935 and 1940. Choosing the 
Greek text of Westcott and Hort as the collating base, Legg supplied 
for Mark (1935) and for Matthew (1940) an enormous thesaurus of 
variant readings of Greek manuscripts, early versions, and patristic 
quotations. It is regrettable that Legg did not indicate in every case the 
edition of the versions and fathers on which he relied. He has been 
criticized also for incomplete citation of evidence as well as occa- 
sional errors." 42 Despite such justifiable criticisms of faults that arise 
chiefly from the ambitious scope of Legg's project — a project that 

42. See, e.g., E. C. Colwell, Classical Philology, xxxiii (1938), pp. 112-15; 
A. Souter, Hxpository Times, liii (1941), pp. 169 ff.; G. D. Kilpatrick, Journal 
of Theological Studies, xliii (1942), pp. 30-4; T. W. Manson, Journal oj "Theo- 
logical Studies, xliii (1942), pp. 83-92. 

192 The Text of the New Testament 

probably exceeded the capacity of any single scholar to accomplish — 
these two volumes present an extraordinary amount of textual infor- 
mation, surpassing any previous apparatus for Matthew and Mark. 43 

Shortly after the publication of the New Testament of the New 
English Bible (1961), requests were received by the Oxford and 
Cambridge University Presses to issue an edition of the Greek text 
that inferentially lies behind the new English version. R. V. G. Tasker, 
a member of the panel of translators of the New English Bible (NEB), 
was entrusted with the task of preparing the edition, which was pub- 
lished in 1964. In an appendix, Tasker cites manuscript evidence 44 
for about 270 sets of variant readings that are represented in the 
margin of the NEB. 45 

In 1966, after a decade of work by an international committee, 46 
five bible societies 47 published an edition of the Greek New 
Testament designed for the use of Bible translators and students 
(Fig. 24). The textual apparatus, which provided a relatively full cita- 
tion of manuscript evidence, included about 1,440 sets of variant 
readings, chosen especially for their exegetical significance. There 

43. In 1949, an international project was launched in order 10 pool the 
resources of textual scholars in Britain and America for the production of 

a comprehensive apparatus criticus of the Greek New Testament. Work 
began on the apparatus for the Gospel of Luke. For a description of the 
initial aims of the project, see M. M. Parvis in Crozer Quarterly, xxvii 
(1950), pp. 301-8. For an account of the continuation of this international 
project, see pp. 248-9. 

44. The very important Bodmer papyri, however, are conspicuous by 
their absence, for most of them were published after the NEB panel had 
completed its work. 

45. For an evaluation of Tasker's edition, reference may be made to 
T. Gaumer, "An Examination of Some Western Textual Variants Adopted in 
the Greek Text of the New English Bible," The Bible Translator, xvi (1965), 
pp. 184-9; B. M. Metzger, Historical and Literary Studies, Pagan, Jewish, 
and Christian (Grand Rapids, MI, 1968), pp. 160-2. 

46. The members of the committee were Kurt Aland of .Miinster, 
Matthew Black of St. Andrews, Allen Wikgren of Chicago, and Bruce 
Metzger of Princeton. During the first 4 years of its work, the committee 
also included the Estonian scholar Arthur Voobus. For the preparation of 
the second edition, which appeared in 1968, the committee was enlarged 
by the addition of Carlo M. Martini, S. J., of Rome. 

47. They are the American, British and Foreign, Dutch, Scottish, and 
Wurttemberg Societies. In 1967, the edition was published under the aus- 
pices of the United Bible Societies. 

The Modern Critical Period 


TL^u "he Eduorial Committee preparing the United Bible Societies' 

Creek New Testament, 2nd and 3rd editions. From right to left: Carlo M. 
Martini, S.J., Kurt Aland, Allen Wikgren, Bruce M. Metzger, and Matthew 
Black (also Klaus Junack, Aland's assistant). 

was also a punctuation apparatus that cited meaningful differences 
in about 600 passages, drawn from five editions of the Greek New 
Testament and from ten translations in English, French, and German. 
A companion volume, entitled A Textual Commentary on the Greek 
New Testament, was drawn up on behalf of the committee by B. M. 
Metzger and published in 1971 by the United Bible Societies. Here, 
the reader can find a succinct account of the committee's reasons for 
either adopting or rejecting this or that variant reading. The com- 
mentary also discusses (beyond the 1,440 sets of variant readings) 
600 other textual problems throughout the New Testament but 
chiefly in the Acts of the Apostles. 

In 1983, the United Bible Societies issued The Greek New Testa- 
ment in what was identified on the title page as the "third edition 
(corrected)." Embodying the work of the staff of the Miinster Insti- 
tute of New Testament Textual Research, under the supervision 
of Klat. s Junack, various corrections were made in the apparatus. 
Likewise, changes were introduced throughout the New Testament, 
conforming the punctuation of the Greek text to that of the Nestle- 
Aland 26th edition, resulting in the introduction of the Continental 
(or Teutonic) tradition of punctuation in place of that reflecting the 
British (Westcott-Hort) tradition. 

194 The Text of the New Testament 

Meanwhile, plans had already been made for the fourth edition 
of the United Bible Societies' Greek New Testament. In 1981, at a 
meeting of the five members of the editorial committee (the places 
left vacant by the retirement of Matthew Black and Allen Wikgren 
were filled by Barbara Aland and John Karavidopoulos of Salonica), 
decisions were made to introduce into the apparatus 284 additional 
sets of variant readings for passages of exegetical importance. Fur- 
thermore, though no change was voted to alter the wording of the 
scriptural text, it was agreed that in some cases a modification 
needed to be made in the assignment of the categories of A, B, C, 
and D relating to the certainty of readings adopted in the text. In a 
number of instances, a C evaluation was raised to the level of B and, 
in a smaller number of instances, a B evaluation was raised to A. It 
was natural that one or both of the new members of the committee 
would favor a certain number of changes, but a growing sense of 
certainty was felt by the three continuing members. 

Eventually, in 1993, the fourth revised edition of The Greek New 
Testament was published by the United Bible Societies. In place of 
the earlier Punctuation Apparatus, the edition contains a Discourse 
Segmentation Apparatus, the work of Roger L. Omanson, a United 
Bible Societies' translations consultant. 

The preceding survey of the more important printed editions of 
the Greek New Testament has referred to only a relatively small pro- 
portion of the total number. No one knows exactly how many sepa- 
rate editions of the Greek Testament have come from the press since 
1514, but it is undoubtedly a very great number indeed. Eduard 
Reuss of Strasbourg, who published a description of editions issued 
up to 1869, was able to enumerate 584 separate ones. 48 If one adds 
the re-editions, the variant editions, and some doubtful editions that 
Reuss mentions in part, the number amounts to 853- Furthermore, 
since Reuss's list is not complete for the more recent period that he 
covers and many editions have appeared since he published his vol- 
ume, it is altogether probable that the 1,000 mark was passed early 
in the twentieth century. 

48. Eduard Reuss, Bibliotheca Novi Testamenti graeci cuius editiones 
ab initio typographiae ad nostrum aetatem impressas quoquot reperiri 
potuerant (Brunsvigae, 1872). For additions to Reuss's list, see Isaac H. 
Hall in Philip Schaff, A Companion to the Greek Testament and the English 
Version, 3rd ed. (New York, 1889), "Supplementary List of Editions, 1871 to 
1882" pp. 520-2. 



The Application of Textual 
Criticism to the Text of the 

New Testament 


The Origins of Textual Criticism 
as a Scholarly Discipline 

Like so many disciplines that we take for granted in our Western 
culture, textual criticism originated among the Greeks. Its rise and 
development were connected with the Homeric epics. Because the 
rhapsodists who recited portions of the Iliad and the Odyssey in 
public would occasionally alter the text to suit the special occasion 
or their own notion of an effective arrangement, there were many 
versions current even in very early times. There subsequently arose 
several "city editions" of Homer, namely, those which presumably 
were preserved by civic authority in various centers (traditionally, 
seven centers) and from which private copies were made. Other 
special texts were made by Theagenes of Regium, Stesimbrotus of 
Thasos (c. 450 B.C.), and Aristotle, who prepared a version for his 
pupil Alexander the Great, usually called rj ex tov vdQOrjxog from 
the case in which it was kept (Plutarch, Life of Alexander 8). 

A more scientific criticism of the text of Homer was developed in 
the Hellenistic Age. This critical study was pursued at the famed library 
in Alexandria, which was reputed to have about 600,000 volumes 1 
and for which, according to tradition, the Greek translation of the 

1. See Edward A. Parsons, The Alexandrian Library, Glory of the 
Hellenic World: Its Rise, Antiquities, and Destructions (New York, 1952); 
J. Harold Ellens, The Ancient Library of Alexandria and Early Christian 
Theological Development ( Occasional Papers, Institute for Antiquity and 
Christianity, 29; Claremont, 1993). 


198 The Text of the New Testament 

Old Testament, the so-called Septuagint, was made. The early directors 
of the library sought to provide ever more accurate editions of the 
Homeric poems. Shortly before 274 B.C., the first of these scholarly 
librarians, Zenodotus of Ephesus (c. 325-c. 240 B.C.), made a com- 
parison of many manuscripts in order to restore the original text of 
both the Iliad and the Odyssey? The corrections that Zenodotus made 
in the text of Homer were of four kinds: (1 ) he eliminated verses that 
he regarded as spurious, (2) he marked others as doubtful but left 
them in his edition, (3) he transposed the order of verses, and (4) he 
introduced new readings not generally current. 

One of the subsequent directors of the library was Aristophanes 
of Byzantium (c. 257-c. 180 B.C.), perhaps the most distinguished 
philologist of Greek antiquity, to whom is ascribed the invention of 
the Greek accent marks as well as other diacritical signs. In his edi- 
tion of the Iliad and the Odyssey, Aristophanes employed a variety of 
critical symbols to indicate his opinion of the state of the text thus 
marked. His greatest pupil was Aristarchus of Samothrace (c. 220- 
c. 144 B.C.), who, becoming his successor at the library, edited the 
works of half a dozen Greek authors and published two critical edi- 
tions of the Homeric poems, supplementing the number of critical 
symbols that his predecessors had used. 

Thus, there was a fairly well-developed scholarly discipline of 
textual and literary criticism in antiquity, localized chiefly at Alexandria 
and directed primarily toward the epics of Homer. It is common 
knowledge that Philo Judaeus and many Church fathers, influenced 
by the philological scholarship current at Alexandria, utilized in their 
interpretation of the Scriptures the methods of allegorical exegesis 
that had been applied to certain stories of the gods and goddesses 
included in the Homeric cycle. It is less widely appreciated — indeed, 
the question has seldom been raised — how far the methods of textual 
criticism current at Alexandria were adopted by scholars in the 
Church and applied to the text of the New Testament. The following 
is a brief summary of what can be learned from patristic sources 
relating to this subject. 

2. See Martin L. West, "The Textual Criticism and Editing of Homer," 
in Editing Texts, ed. by Glenn W. Most (Aporemata, Kritische Studien zur 
Philologie-geschichte, 2; Gottingen, 1998), pp. 94-110: "By 1990 the num- 
ber of the manuscript tradition had reached 703 for the Iliad and 236 for 
the Odyssey (p. 102). 

The Origins of Textual Criticism as a Scholarly Discipline 199 

Ironically enough, the earliest efforts 3 to ascertain the original 
text of the New Testament seem to have been made by those who 
were excommunicated as heretics by the authoritarian bishop of 
Rome, Pope Victor I (served as pope c. A.D. 187-98). It appears that 
a learned leather merchant (oxvrevg) named Theodotus, lately 
come from Byzantium to Rome, had been stung by certain criticisms 
that Galen, the famous Greek physician, had leveled against the 
philosophical naivete of many Christians.' 1 In an attempt to intro- 
duce improvements in the methodology of scriptural interpretation, 
Theodotus and his followers seem to have undertaken a critical 
recension of the biblical text. Eusebius preserves a large excerpt of 
an almost contemporary pamphlet by an anonymous author di- 
rected against these philosophically minded Christians. 3 According 
to this author, the Theodotians deserved to be condemned on three 
scores: (1) they were engrossed in the study of logic, mathematics, 
and empirical science ("Some of them, in fact, study the geometry of 
Euclid, and admire Aristotle and Theophrastus; and Galen perhaps 
is even worshipped by some of them"); (2) rejecting allegorizing, 
they practiced strict grammatical exegesis; and (3) they applied tex- 
tual criticism to the Septuagint and the Greek New Testament: 

They did not fear to lay hands on the divine Scriptures, alleging that 
they had critically revised [dtcoodcDxevai] them. . . . For they cannot 
deny that this audacious act is their own, seeing that the copies are 
written in their own hand, and they did not receive the Scriptures in 
this condition from their teachers, nor can they show any copies from 
which they made their emendations I6el^ai avriyga^a oOev avxa 
/lersygdipavro fifj exwoiv]? 

Unfortunately, nothing more is known of this early effort at textual 

3. The alterations made by Marcion in the New Testament were 
motivated by doctrinal considerations rather than by an interest in textual 

4. See R. Walzer, Galen on Jews and Christians (Oxford, 1949), pp. 75 ff. 

5. Eusebius, Hist. Eccl. v.xxviii. 13-19 (the excerpt may be from 
Hippolytus of Rome's Little Labyrinth) . 

6. For a discussion of this passage, see Hermann Schone, "Ein 
Einbruch der antiken Logik und Textkritik in die altchristliche Theologie," 
in Pisciculi: Studien zur Religion und Kultur des Altertums; Franz Joseph 
Dolger . . . dargeboten , . . (Munster, 1939), pp. 252-65. 

200 The Text of the New Testament 

Not long after the Theodotians had been excommunicated, one 
of the most assiduous and erudite scholars of his age, Origen of 
Alexandria and Caesarea, began a text-critical study of the entire Old 
Testament in Hebrew and in several Greek translations. His resulting 
Hexapla, which must have required many years of the most painstak- 
ing labor, was a monumental tool that many patristic scholars consulted 
in the famed library of Pamphilus at Caesarea, until its destruction in 
the seventh century during the Islamic conquest of the Near East. 

The question of whether Origen ever attempted to edit a critical 
text of the New Testament has been answered quite diversely by 
modern scholars; 7 it seems probable to most investigators that he did 
not extend his textual efforts to preparing a formal edition of the 
New Testament. At the same time, in all his writings and particularly 
in his exegetical treatises, Origen reveals a certain solicitude for critical 
details in the biblical text. He complains that 

the differences among the manuscripts [of the Gospels] have become 
great, either through the negligence of some copyists or through the 
perverse audacity of others; they either neglect to check over what they 
have transcribed, or, in the process of checking, they lengthen or 
shorten, as they please. 8 

Besides making comments of a general nature about the text, Origen 
sought out information (though he did not always utilize this infor- 
mation) concerning variant readings in Greek manuscripts of the 
New Testament. He observes, for example, that in Matthew's account 
(18.1) of the disciples' question as to who is the greatest in the king- 
dom of heaven, according to some of the manuscripts the Evangelist 
prefixed the phrase iv ixeivfl 177 &gg., whereas according to others 
the expression iv exeivrf xfj tffiega appears. 9 Similarly, Origen notices 
the two readings in Heb. 2.9, "apart from God" (xcogigOEOv) and "by 
the grace of God" (%&qiti deov) but is not interested in deciding 
between them, for he finds spiritual significance in both. 

7. For a summary of these opinions, see B. M. Metzger, "Explicit 
References in the Works of Origen to Variant Readings in New Testament 
Manuscripts," in Biblical and Patristic Studies in Memory of Robert Pierce 
Casey, ed. by J. N. Birdsall and R. W. Thomson (Freiburg, 1963), pp. 78-95. 

8. Commentary on Matthew xv. 14 {Die griechiscben christlicben 
Schriftsteller, Origenes, x.387. 28-388.7, ed. Klostermann). 

9. Comm. on Matt. xiii. 14 (G.C.S., Origenes, x. 213. 21 ff, ed. 

The Origins of Textual Criticism as a Scholarly Discipline 201 

At other times Origen declared his preference among variant 
readings, but often his choice appears to be based on considerations 
other than those of a purely textual nature. Thus, when he dismisses 
the reading "Jesus Barabbas" in favor of simply "Barabbas" (Matt. 
27.16-17), he does so because he thinks that the name "J esus " was 
never applied to evil-doers. 10 Again, Origen's well-known prefer- 
ence for the reading "Bethabara" instead of "Bethany" as the place of 
John's baptizing (John 1.28) was adopted on geographical and ety- 
mological grounds, 11 and the same reasons dictated his preference 
for "Gergesa" rather than "Gerasa" or "Gadara" as the name of the 
place where the demons entered the herd of swine. 12 In a different 
category are instances where, because of some exegetical difficulty, 
Origen suggests that perhaps all of the manuscripts existing in his 
day may have become corrupt. 13 

Judged according to modern standards, St. Jerome (c. 347-420) 
was a more sagacious textual critic than Origen, well aware of the 
varieties of error that arise in the transcription of manuscripts. He 
refers, for example, to the possibility of confusion of similar letters, 
confusion of abbreviations, accidents involving dittography and hap- 
lography, the metathesis of letters, assimilation, transpositions, and 
deliberate emendations by scribes. 14 Several explicit references will 
indicate his interest in text-critical details. In the preface to his revi- 
sion of the Latin Gospels, addressed to Pope Damasus, who had re- 
quested that he undertake the work, Jerome declares that for the tex- 
tual basis of the revision he relied upon older Greek manuscripts. 
Again, in his letter to Minervius and Alexander, 15 two monks at 
Toulouse who had written to Jerome asking him to explain certain 
passages in Scripture, Jerome discusses several forms of the text of 1 

10. In Matt. Comm. ser. 121 (G.C.S., Origenes, xi.2.255, 24 ff., ed. 

11. Commentary on Joan, vi.40 (24) (G.C.S., Origenes, iv.149.12 ff., 
ed. Preuschen). 

12. Comm. on Joan, vi.41 (24) (G.C.S., Origenes, iv.150.3 ff., ed. 

13. For examples, see Metzger, op. cit. 

14. For one or more examples from the works of Jerome that illustrate 
each of these categories, see K. K. Hulley, "Principles of Textual Criticism 
Known to St. Jerome," Harvard Studies in Classical Philology, lv (1944), 
pp. 87-109. 

15. Epist. 119 (Migne, Patrologia Latina, xxii, 966 ff.). 

202 The Text of the New Testament 

Cor. 15.51 ("We shall not all sleep, but we shall all be changed"). He 
indicates that he prefers the reading "We shall all sleep, but we shall 
not all be changed." In his Dialogue Against the Pelagians, 16 Jerome 
states that in certain copies, especially in Greek codices, an extensive 
addition was to be found at the close of the Gospel according to 
Mark. Jerome does not tell us where he found these manuscripts, 
and no such copy was known until the twentieth century, when the 
passage turned up in a Greek manuscript which Charles L. Freer of 
Detroit had bought from an Arab dealer in Gizeh near Cairo (for the 
translation of this addition, see p. 81). 

Although primarily a theologian, St. Augustine (354-430) showed 
on occasion a keen critical judgment in textual problems. Thus, when 
considering the difficulty that Matthew (27.9) attributes a quotation 
to Jeremiah that actually appears in Zechariah, Augustine suggests 
that one should 

first take notice of the fact that this ascription of the passage to Jere- 
miah is not contained in all the manuscripts of the Gospels, and that 
some of them state simply that it was spoken "by the prophet". It is 
possible, therefore, to affirm that those manuscripts deserve rather to 
be followed that do not contain the name of Jeremiah. For these 
words were certainly spoken by a prophet, only that prophet was 

With commendable candor, however, Augustine declares that he is 
not altogether satisfied with this explanation, because "a majority of 
manuscripts contain the name of Jeremiah, and those who have 
studied the Gospel with more than usual care in the Greek copies re- 
port that they have found it to stand so in the more ancient Greek 
exemplars." Thereupon, Augustine virtually enunciates the critical 
canon that the more difficult reading is to be preferred: 

I look also to this further consideration, namely that there was no rea- 
son why this name should have been added [subsequently to the true 
text] and a corruption thus created; whereas there was certainly an in- 
telligible reason for erasing the name from so many of the manuscripts. 
For presumptuous inexperience (audax imperitid) might readily have 
done that, when perplexed with the problem presented by the circum- 
stance that this passage cannot be found in Jeremiah. 17 

16. Dialog, contra Pelagianos, ii. 15 (Migne, P.L., xxiii, 576). 

17. De consensu Evangel, iii.7.29 (Migne, P.L., xxxiv, 1174 f.). 

The Origins of Textual Criticism as a Scholarly Discipline 203 

On another occasion, Augustine suggests that preference should 
be given to readings that are current in important sees, thus 
anticipating B. H. Streeter's theory of "local texts" (see pp. 214-6). 
He writes: "If the books of the New Testament are confusing in the 
variety of their Latin translations, they should certainly give place to 
the Greek versions, especially to those that are found in churches of 
greater learning and research." 18 

During the Middle Ages, when knowledge of Greek was at a low 
ebb, text-critical efforts were now and then directed toward the pu- 
rification of Jerome's Vulgate text. It was perhaps to be expected that 
this version, besides being corrupted with the usual types of error in- 
cident to all transcription, would once again incorporate certain Old 
Latin readings which Jerome had eliminated from his text. (For some 
of these attempts to purify Jerome's text, see p. 106.) The writings 
of such authors as Gilbert of Porree and Peter Lombard contain 
sporadic comments, reflecting information derived from Jerome and 
Augustine, regarding the Greek lying behind such and such Latin 
renderings. 19 

At the time of the Renaissance and with the spread of the knowl- 
edge of ancient Greek, scholars began to correct the Latin Vulgate 
by the original Greek. In their biblical annotations, Erasmus and 
Beza not infrequently refer to variant readings in Greek manuscripts. 
As mentioned in Chapter 3, the first English Bible to contain the 
translation of variant readings from Greek manuscripts (including 
Codex Bezae) was the Geneva Bible of 1560, prepared by William 
Whittingham and other English exiles residing at Geneva. This Bible 
has, for example, the negative Golden Rule in the margin opposite 
Acts 15.29, as well as a translation of the Western addition at Acts 
19.9 that Paul preached daily at the school of Tyrannus "from the 
fifth to the tenth hour." 20 

18. De doctrina Christiana, ii. 15.22 (Migne, P.L., xxxiv, 46): "apud 
ecclesias doctiores et diligentiores." 

19. See Arthur Landgraf, "Zur Methode der biblischen Textkritik im 12. 
Jahrhundert," Biblica, x (1929), pp. 445-74; M.-J- Lagrange, "La critique 
textuelle avant le concile de Trente," Revue Thomiste, xxxix.2 (1934-5), 
pp. 400-9 (Lagrange's La Critique textuelle; ii, La critique rationnelle 
[Paris, 19351, pp. 294-301). 

20. The equivalent according to the modern reckoning of time is from 

1 1 a.m. to 4 p.m. For a score of other variant readings given in the margins of the 
Geneva Bible, see Metzger, "The Influence of Codex Bezae," p. 151, n. 34. 

204 Tin; Text of the New Testament 

The first scholar to make any use of all three classes of evidence 
for the text of the New Testament — that is, Greek manuscripts, the 
early versions, and quotations from the fathers — was probably Francis 
Lucas of Bruges (Brugensis) in his Notationes in sacra Biblia, quibus 
variantia . . . discuthmtur (Antwerp, 1580). Toward the close of the 
seventeenth century, the scientific foundations of New Testament 
criticism were laid in four monumental publications of Richard 
Simon (1638-1712), a French Catholic scholar far ahead of his day 
in biblical research. The volumes are entitled Histoire critique du 
texte du Nouveau Testament (Rotterdam, 1689); English trans., Critical 
History of the Text of the New Testament, 2 parts (London, 1689); His- 
toire critique des versions du Nouveau Testament (Rotterdam, 1690); 
English trans., Critical History of the Versions of the New Testament 
(London, 1692); Histoire critique des principaux commentateurs du 
Nouveau Testament, 2 parts (Rotterdam, 1693); and Nouvelles obser- 
vations sur le texte et les versions du Nouveau Testament (Paris, 
1695). Disregarding the traditional and dogmatic presuppositions of 
his age, Simon examined critically the text of the Bible as a piece of 
literature. His works are full of acute observation and reasoning and 
anticipate in detail many of the conclusions of scholars two and 
three centuries later. 


Modern Methods of 
Textual Criticism 

I. The Classical Method of 
Textual Criticism 

The method of textual criticism that has been generally practised 
by editors of classical Greek and Latin texts involves two main 
processes, recension and emendation. Recension is the selection, 
after examination of all available material, of the most trustworthy 
evidence on which to base a text. Emendation is the attempt to elim- 
inate the errors that are found even in the best manuscripts. 1 

The classical method of textual criticism arose during and after the 
Renaissance, when attention was drawn to spurious papal decretals 

1. For a concise account of these two basic processes, see Paul 
Maas, Textual Criticism (Oxford, 1958). Other useful discussions of the 
classical procedures of textual criticism include the following: R. C. Jebb 
in A Companion to Greek Studies, ed. by Leonard Whibley (Cambridge, 
1906), pp. 610-23; J. P. Postgate in A Companion to Latin Studies, ed. by 
J. E. Sandys (Cambridge, 1910), pp. 791-805; L. Havet, Manuel de 
critique verbale appliquee aux textes latins (Paris, 1911); F. W. Hall, A 
Companion to Classical Texts (Oxford, 1913), pp. 108-98; H. Kantorowicz, 
Einfilhrung in die Textkritik (Leipzig, 1921); L. Bieler, "The Grammarian's 
Craft," Folia: Studies in the Christian Perpetuation of the Classics, x (1956), 
pp. 3-42. 


206 The Text of the New Testament 

and questions were raised regarding falsifications in Church history 
and in the credentials of certain religious orders. The critical acumen 
of scholars was sharpened likewise by the large number of forged 
texts that began to appear; for example, a single forger, Giovanni 
Nanni (alias Joannes Annius, 1432-1502), a Dominican monk of 
Viterbo, put forth 17 spurious treatises attributed by him to ancient 
Greek and Latin authors. 2 

A more critical spirit in dealing with ecclesiastical documents 
found expression during the sixteenth century in the work of 
Matthias Flacius and the group of Lutheran scholars known as the 
Magdeburg Centuriators, who were the first to write the history of 
the Church from a Protestant point of view. In 1675, the Jesuit 
scholar Daniel Papebroch aroused the hostility of the Benedictines 
by denying the authenticity of documents constituting the credentials 
of certain Benedictine monasteries. The learned Benedictine monks 
at St. Maur took up the challenge by founding the science of pale- 
ography, which is the classification of manuscripts according to their 
age in the light of their handwriting and other indications. The first 
treatise to deal with the Latin paleography of official documents 
was the monumental work of the Maurist Jean Mabillon (1632-1707) 
entitled De re diplomatica (Paris, 1681). The science was extended to 
Greek manuscripts by another Benedictine, Bernard de Montfaucon 
(1655-1741) in his Palaeographica graeca (Paris, 1708). 

The application of critical methods to the editing of classical 
texts was developed principally by three German scholars, Friedrich 
Wolf (1759-1824), one of the founders of classical philology; 
Immanuel Bekker (1785-1871); and Karl Lachmann (1793-1851). 
Bekker devoted his long life to the preparation of critical editions of 
Greek texts. The transfer of many manuscripts to public libraries as 
a result of the upheaval following the French Revolution provided 
the opportunity for extensive collation of manuscripts older than 
those that had previously been generally available. Bekker collated 
some 400 manuscripts, grouped existing manuscripts of an author 

2. Annius' volume De Commentariis antiquitatum (Rome, 1498) 
includes the allegedly rediscovered writings of Xenophon, Berosus, Cato, 
Antoninus Pius, Manetho, Philo Judaeus, Caius Sempronius, and Myrsilus, 
all of which, Annius declared, he had found buried in the ground. See 
"Nanni" in Pierre Bayle's Dictionaire historique et critique and "Annius" in 
J. S. Ersch and J. G. Gruber's Allgemeine Encyclopadie; James Farrer, 
Literary Forgeries (London, 1907), pp. 67-81. 

Modern Methods of Textual Criticism 


into families where one was derived from another, and published 
60 volumes of improved editions of Greek authors. As was mentioned 
earlier (p. 170), Lachmann went further than Bekker, showing how, 
by comparison of manuscripts, it is possible to draw inferences as to 
their lost ancestors or archetypes, their condition, and even their 

The basic principle that underlies the process of constructing a 
stemma, or family tree, of manuscripts is that, apart from accident, 
identity of reading implies identity of origin. By way of example, 
suppose that there are seven manuscripts of an ancient book and 
that in a certain paragraph three of them agree in lacking a sentence 
that is present in the other four manuscripts. From this circumstance 
we would deduce either that a common ancestor of the three had 
omitted the sentence or that an ancestor of the four had added it. 
Suppose, moreover, that we find that the seven manuscripts fre- 
quently range themselves so that one of them (which we may desig- 
nate A) stands apart, showing no great similarity to any of the other 
six, while B, C, and D, on the one hand, and E, F, and G, on the 
other hand, greatly resemble each other, though differing somewhat 
from the rest. We can express this by saying that B, C, and D form a 
family, descended from a hypothetical common ancestor that we 
may call X, and that E, F, and G form another family, descended from 
a hypothetical ancestor that we may call Y. The readings of X that 
can be deduced by comparing those of B, C, and D will be of a 
higher antiquity and of greater authority than any of the readings in 
B, C, or D taken singly; and the same may be said for the readings 
of Y when compared with those in E, F, and G. Indeed, it is possible 
to go further: we may compare the readings of X and Y with each 
other and with those of A and thus deduce the readings of a still 
more remote ancestor that we may call Z, the hypothetical archetype 
of all the manuscripts. Thus, the pedigree of all ten manuscripts (the 
seven extant and the three hypothetical) would be as follows: 

208 The Text of the New Testament 

It follows that because B, C, and D may agree in a given reading 
against A, such a reading is not three times more likely to be correct 
than the reading in A. In fact, it is obvious that, other things being 
equal, there is a fifty-fifty chance that either of the two readings may 
be correct, for where B, C, and D agree they represent a manuscript 
(X) which is as far removed from the archetype (Z) as is A. Thus, in- 
stead of merely counting the number of manuscripts supporting a 
given reading, the editor must weigh their significance in accordance 
with their mutual relations. 

Often, however, difficulties hinder the construction of a stemma 
of manuscripts. The simple example given above assumes that the 
different lines of descent have remained independent of one another. 
But a disturbing element enters when mixture has occurred, that is, 
when a copyist has had two or more manuscripts before him and has 
followed sometimes one and sometimes the other or, as sometimes 
happened, when a scribe copied a manuscript from one exemplar 
and corrected it against another. To the extent that manuscripts have 
a "mixed" ancestry, the genealogical relations among them become 
progressively more complex and obscure to the investigator. 3 

Noteworthy among scholars of the Greek and Latin classics who 
have given attention to the methodology and practice of textual crit- 
icism is Martin L. West, Fellow of University College, Oxford. Well 
known as the editor of Greek poetic texts, West was invited by the 
celebrated publishers of classical texts B. G. Teubner of Germany to 
write a manual for prospective editors of classical texts that would 
take the place of Maas's Textkritik (see p. 211, n. 11). The latter had 
emphasized the stemmatic aspect of textual analysis and treated con- 
tamination (i.e., the presence of readings in a manuscript derived 
from two or more earlier manuscripts) as a regrettable deviation 
about which nothing can be done. In trying to redress the balance, 
West discusses contaminated tradition in specimen passages, ranging 

3- Genealogical mixture is widespread in the textual transmission of 
most Latin patristic literature; see the sane and balanced account in Maurice 
Bevenot, The Tradition of Manuscripts: A Study in the Transmission of 
St. Cyprian's Treatises (Oxford, 1961). I ; or the special problems that 
confront the editor of Greek patristic documents, see Herbert Musurillo, 
"Some Textual Problems in the Editing of the Greek Fathers," Texte und 
Untersuchungen, lxxviii (1961), pp. 85-96. 

Modern Methods of Textual Criticism 209 

from Hesiod to Ovid and Apuleius, and gives the reader practical 
advice in dealing with a variety of textual problems. 4 

Several other scholars have made significant contributions to the 
classical method of textual criticism. The Sather Lectures by E. J. 
Kenney of Peterhouse, Cambridge, on the history of the classical tradi- 
tion are, as would be expected, wide-ranging and urbane. 5 Concen- 
trating on the art of emendation as practiced by earlier scholars, the 
author summarizes his argument by referring to A. L. von Schlozer's 
insistence that method is not everything and that "there is something 
in criticism which cannot be subjected to rule, because there is a 
sense in which every case is a special case." 6 According to Kenney, 
"These are words that merit quotation and remembrance, since they 
enshrine what must be accounted the only completely and universally 
valid principle of textual criticism ever formulated." 

In what he describes as a workbook dealing with textual criti- 
cism, Robert Renehan 7 discusses 82 passages in a broad range of 
classical authors where "something is amiss that needs to be set 
right by the critic." More than once, however, Renehan has to admit 
that certainty is not attainable, and he frequently contents himself 
with merely enumerating the choices among variant readings in 
manuscripts and conjectural emendations proposed by various 
critics. From his examples the student soon learns that when one is 
deciding that a given reading is derived from another, and there- 
fore is to be eliminated, one is doing something similar to what is 
done when deciding that a given manuscript is derived from 

4. M. L. West, Textual Criticism and Editorial Technique Applicable to 
Greek and Latin Texts (Stuttgart, 1973). See also his more recent Studies in 
the Text and Transmission of the Iliad (Munich, 2001). 

5. E. J. Kenney, The Classical Text: Aspects of Editing in the Age of the 
Printed Book (Berkeley, 1974). 

6. Ibid., p. 98. von Schlozer was an eighteenth-century historian at 
Gottingen, who, according to Kenney, relied heavily on methods worked 
out by New Testament scholars. In fact, in another connection, Kenney 
acknowledges, "In the matter of the apparatus criticus, as in more 
fundamental matters, it was New Testament criticism that gave the lead 
[to classical scholarship]," (p. 156). 

7. R. Renehan, Greek Textual Criticism, A Reader (Cambridge, MA, 

210 The Text of the New Testament 

II. Reactions Against Classical 
Textual Criticism 

1. Joseph. Bedier 

During the twentieth century, the genealogical method came 
under attack from several quarters; some scholars rejected it entirely, 
while others restricted its application to a narrow and limited area. 
One of the former is Joseph Bedier, an editor of several medieval 
French documents. It was while preparing his edition of Le lai de 
I 'ombre par Jean Renart (Paris, 1913), so Bedier declared, that he 
became distrustful of the genealogical method, (1) because in prac- 
tice it has almost always resulted in the construction of a tree with 
two branches of witnesses (a circumstance that Bedier viewed with 
cynical suspicion, suggesting that editors have deliberately forced 
the evidence into a stereotyped pattern) and (2) because one can 
often argue well for several different stemmata of classification of 
manuscripts. Bedier's own method was to choose what seems to be 
the best manuscript, on the basis of grammar, sense, and simple and 
regular orthography, and then to use the other manuscripts eclecti- 
cally in correcting sporadic readings in the manuscript accepted as 
primary. 8 

Among New Testament scholars who came under the influence 
of Bedier's skepticism of the value of the genealogical method were 
Leon Vaganay and Ernest Cadman Colwell. The former roundly as- 
serted that "applied to New Testament texts this system is useless." 9 
The latter more cautiously declared that: 

the genealogical method is not of primary importance. ... It can chart 
the history of transmission in an area narrowly limited in time and 
space. . . . But in the larger area where the larger questions are settled, 
it still has to demonstrate its value for the reconstruction of the original 
text of the Greek New Testament. 10 

8. Besides the preface in his 1913 edition, see Bedier's fuller exposi- 
tion in "La Tradition manuscrite du Lai de I'ombre: reflexions sur 1'art 
d'editer les anciens textes," Romania, liv (1928), pp. 161-96, 321-56. 

9. Leon Vaganay, Initiation a la critique textuelle neotestamentaire 
(Paris, 1934), p. 60; English trans., p. 71. 

10. E. C. Colwell, "Genealogical Method: Its Achievements and its 
Limitations," repr. in Ernest C. Colwell, Studies in Methodology in Textual 

Modern Methods of Textual Criticism 211 

In evaluating the justice of Bedier's attack on the genealogical 
method, it ought to be pointed out that a much more innocent 
explanation lies behind the circumstance that almost all stemmata 
result in two branches than the imputation of deliberate suppression 
or distortion of evidence. From the standpoint of mathematics, as 
Maas observed, "We must remind ourselves that of the twenty-two 
types of stemma possible where three witnesses exist, only one has 
three branches." 11 

Bedier's second criticism of Lachmannian methodology pos- 
sesses a semblance of justification only when one treats manuscripts 
as though they were fixed entities, as unchanging as a printed book. 
On the contrary, after a manuscript was copied, it continued to live 
and was subject to modifications — as the numerous erasures, correc- 
tions, additions, glosses, and remarks by readers entered in the mar- 
gins testify. 12 One must therefore take into account what may be 
called the successive "stages of manuscripts" 13 as well as the possi- 
bility of multiple parentage. 

The disconcerting ambiguity arising from the construction of 
equally cogent classifications of the same manuscripts need not result 
in abandoning the Lachmannian method altogether. Faced with a 
number of manuscripts that have a mixed ancestry, the textual critic 
may well adopt a compromise between the purely eclectic method 
advocated by Bedier and the rigid genealogical procedure of classical 
text criticism. Suppose that there are five manuscripts, A, B, C, D, and 
E, the lineage of which is not clear; we cannot say, for example, that 
A, B, and C form one family, descended from a common ancestor, 

Criticism of the New Testament (New Testament Tools and Studies, IX; 
Grand Rapids, MI, 1969), pp. 65-83. See also his chapter, in the same 
volume, "Method in Grouping of New Testament Manuscripts," pp. 1-25. 
For a fuller appreciation of classical approaches to textual criticism by the 
same author, see his chapter, also in the same volume, "Hon Redivivus, A 
Plea and a Program," pp. 148-71. 

11. Maas, Textual Criticism (Oxford, 1958), p. 48. 

12. For a strong statement of this point with respect to the textual 
tradition of the New Testament Gospels, see David C. Parker, 'lhe Living 
Text of the Gospels (Cambridge, 1997). 

13. For a discussion of the implications of such alterations, marking 
successive stages in the history of the transmission of a treatise, see Jean 
Irigoin, "'Stemmas bifides et etats des manuscrits," Revue de philologie, lxxx 
(1954), pp. 211-17. 

212 The Text of the New Testament 

while D and E form another. A comparison discloses, however, that 
certain characteristic readings are common to the group A, B, and C 
but not present in D and E. This evidence shows that, so far as these 
readings are concerned, some manuscript with such characteristic 
readings was one member in the ancestry common to A, B, and C, 
though that ancestry may in other respects be mixed. 

One may conclude, therefore, that, despite the presence of a 
large amount of mixture in the ancestry of New Testament manu- 
scripts, it will be advantageous for the textual critic to search out the 
broad features of more or less closely related groups of manu- 
scripts. 14 Such a process discloses that in general the Koine (or 
Byzantine) text of the New Testament is secondary, being character- 
ized by the features that Hort delineated with classic vividness (see 
pp. 178-9). Moreover, major dislocations common to the members of 
smaller groups of manuscripts prove both the existence of such 
groups and their ultimate derivation from a common archetype that 
had suffered such dislocations (e.g., the position of the pericope de 
adultera after Luke 21.38 in fam. 13). 15 

2. Albert G. Clark 

One of the axioms of classical textual criticism is brevior lectio po- 
tior; that is, the shorter of two readings is probably original. This prin- 
ciple, which has been accepted as generally valid by both classical 
and biblical scholars, was challenged in 1914 by Albert C. Clark in his 
inaugural lecture as Corpus Professor of Latin at Oxford University. 16 
Clark's researches on the manuscripts of Cicero's orations led him to 
believe that accidental omission was a much more common fault than 
deliberate interpolation by scribes. Four years later, Clark published a 

14. See p. 87. 

15. Although the presence of mixture to a greater or lesser extent in 
New Testament manuscripts makes it impossible to draw up precise 
genealogical stemmata, the mixture itself carries with it compensatory 
benefits. As Zuntz pointed out {Classica et Mediaevalia, Hi [1940], p. 24), 
the more widely that cross-fertilization of manuscripts has taken place, 
the more improbable it is that any old reading, true or false, could have 
entirely disappeared. 

16. Albert C. Clark, Recent Developments in Textual Criticism, an 
inaugural lecture delivered before the university on 6 June 1914 (Oxford, 

Modern Methods of Textual Criticism 213 

lengthy treatise, The Descent of Manuscripts (Oxford, 1918), in which 
he showed that many omissions in classical texts involve multiples of 
the number of letters in an average line of script. Of two forms of text, 
one longer and one shorter, the latter can almost always be explained 
as the result of a scribe's omitting one or more lines of his exemplar. 
As Clark put it, "A text is like a traveller who goes from one inn to 
another, losing an article of luggage at each halt." 17 

Clark applied his principle longior lectio potior 'to the text of the 
Gospels and Acts, 18 with the result that the Western form of text, 
being in general the longer text, came off much better than it had at 
the hands of Westcott and Hort. If Hort could see no good in the 
Western text, Clark could see none in the Neutral text, which he re- 
garded as the result of accidental omissions of multiples of lines of 
average length. 

Clark's theory of accidental scribal omissions was criticized on 
several scores by such eminent textual scholars as Sanday, Souter, and 
Kenyon. The latter pointed out, 19 for example, that (1) variation in the 
length of lines in manuscripts makes the method of counting letters 
unreliable except for very short passages; (2) accidental omissions 
would not account for the regular correspondence of the omissions 
with breaks in the sense; (3) most of the variants involve, not omis- 
sion, but differences in wording; and (4) such narrow columns as 
Clark's theory necessitates are exceedingly rare in the early papyri (an 
argument that has become the stronger as more and more early papyri 
with relatively wide columns have come to light). Furthermore, the 
circumstances of the transmission of the Gospel accounts were quite 
different from those of Cicero's Verrine orations. The Church pre- 
served many traditions of the deeds and sayings of Christ that had not 
been included in the Gospels (e.g., John 21.25). It would be natural 
for these to slip into the text of the Gospels, either from the margins 
of other manuscripts or from the living memory of the Church. 

In a subsequent study of the Western text of the Acts, Clark 
returned to the controversy. 20 This time, he practically abandoned 

17. A. C. Clarke, Journal of "theological Studies, xvi (1915), p. 233. 

18. Albert C. Clark, The Primitive Text of the Gospels and Acts (Oxford, 

19. F. C. Kenyon, The Text of the Greek Bible (London, 1937), p. 231; 
idem, "The Western Text in the Gospels and the Acts," Proceedings of the 
British Academy, xxiv (1938), pp. 287-315. 

20. AtbenC. Clark, The Acts of the Apostles (Oxford, 1933). 

214 The Text or the New Testament 

the theory of accidental omission and revived the suggestion pro- 
posed in the seventeenth century by Jean leclerc, that Luke had him- 
self produced two editions of Acts. This hypothesis shifts the inquiry 
from scribal transmission to deliberate editorial alteration on the part 
of the author or editor, and its validity must be tested on grounds 
other than those of textual criticism alone. The only comment that 
needs to be made here is that a comparison of the trends in the tex- 
tual criticism of the Iliad and the Mahdbhdrata, two great national 
epics the transmission of which reveals certain parallels to the trans- 
mission of the Gospels, is instructive for the New Testament scholar. 
Textual critics of both these corpora of quasi-religious literature are 
convinced that they are growing texts and that no scribe deliberately 
excised any considerable portion of either poem. 21 

More recently, the case against the criterion brevior lectio potior, 
at least for the earliest New Testament witnesses, has been taken up 
by James Royse, who, on the basis of a careful study of the papyri, 
has concluded that in fact the opposite scribal tendency appears to 
hold, that is, that the scribes of our surviving papyri were more likely 
to omit portions of the text rather than add to it. 22 

III. Local Texts and Ancient Editions: 
Burnett Hillman Streeter 

One New Testament scholar who made good use of classical 
methods of textual criticism was B. H. Streeter, who in 1924 pub- 
lished The Four Gospels, A Study of Origins, in which solid scholar- 
ship is combined with a fertile imagination and an engaging literary 

21. For a survey of the trends of textual scholarship in the editing 

of the Iliad and the Mahdbhdrata, reference may be made to B. M. Metzger, 
Chapters in the History of New Testament Textual Criticism (Grand Rapids, 
MI, 1963), pp. 142-54. The conclusion mentioned above does not detract 
from the principle that before a book has become sacred, careless copying is 
much more likely to result in omissions than additions. 

22. James R. Royse, "Scribal Habits in Early Greek New Testament 
Papyri," (Diss., Berkeley, CA: Graduate Theological Union, 1981); see 
also idem, "Scribal Tendencies in the Transmission of the Text of the 
New Testament" in The Text of the New Testament in Contemporary 
Research: Essays on the Status Quaestionis, ed. by Bart D. Ehrman and 
Michael W. Holmes (Grand Rapids, MI, 1995) pp. 239-52. 

Modern Methods of Textual Criticism 215 

style. Building on Westcott and Hoit's classic work, Streeter refined 
their methodology in light of the acquisition of new manuscript evi- 
dence since 1881. Adopting an idea that J. L. Hug had first devel- 
oped, Streeter emphasized the importance of isolating the forms of 
text that were current at the great centers of ancient Christianity. By 
means of evidence derived from quotations in the writings of early 
Church fathers, he isolated and identified the characteristic forms of 
New Testament text that had developed at the principal sees of the 
ancient Church. By about A.D. 200, these local texts had reached 
(so Streeter believed) their maximum divergence, a divergence that 
is reflected in the earliest Syriac, Latin, and Coptic versions. It is 
probable that the oldest forms of these three versions were derived, 
respectively, from the Greek texts current in Antioch, Rome, and 

Besides these three forms of text, Streeter's analysis of the evi- 
dence of Codex Koridethi (<9) and some of the writings of Origen 
and Eusebius led him to postulate the existence of a so-called Cae- 
sarean text of the Gospels, to which fam. 1 and fam. 1 3 also belong. 
Streeter combined into one text type, which he designated the 
Alexandrian, the witnesses that Westcott and Hon" had assigned to 
their Neutral and Alexandrian groups. He agreed with Westcott and 
Hort that the Syrian text, which he renamed the Byzantine text, arose 
during the fourth century through the recensional activity of Lucian 
of Antioch and was adopted by about 380 at Constantinople. This 
text became the prevailing ecclesiastical form of the New Testament 
throughout the Greek-speaking world and eventually constituted the 
basis of the Textus Receptus. Therefore, readings later than the fifth 
century, Streeter argued, can be ignored except when they differ 
from the prevailing Byzantine text. On the other hand, because it is 
possible that an ancient form of text may have been preserved at a 
relatively late date in a locality cut off from the main stream of Chris- 
tianity, the precedence of manuscripts depends not so much on their 
age as on their pedigree. 

The relationship between the local texts used in the five 
Churches of Alexandria, Caesarea, Antioch, Italy and Gaul, and 
Carthage- — stands in a graded series corresponding to their geo- 
graphical propinquity around the eastern Mediterranean world. 
Streeter sets forth this point as follows: 

Each member of the series has many readings peculiar to itself, hut 
each is related to its next-door neighbour far more closely than to re- 
moter members of the series. Thus B (Alexandria) has much in 


The Text of the New Testament 

common with Jam. (Caesarea);/«w. shares many striking readings 
with Syr. S. (Antioch); Syr. S. in turn has contacts with T> b a 
(Italy-Gaul); and, following round the circle to the point from which 
we started, k (Carthage) is in a sense a half-way house between D b a 
and B (Alexandria again). 23 

Charts 1 and 2 reproduce a stemma and a chart from Streeter's 
volume, showing the relationship of the several local texts and the 
chief witnesses that support each. Some of the practical conclusions 
that follow from the acceptance of Streeter's theory of local texts 
include the following, which are set forth in his own words. 

1. The textual critic, in weighing the amount of external evidence in favor 
of any reading, should consider primarily, not the number or the age 
of the manuscripts which support it, but the number and geographical 
distribution of the ancient local texts in which it can be traced. 

2. It follows that manuscripts should be cited, not in alphabetical or nu- 
merical order, but in groups corresponding to the local texts that they 
represent. When at least three of the leading representatives of any 

Original Autographs 




CL 33 Boh 

Revised Text of I.ucian c.A.D.310 

Byzantine (or "Standard") Text. (A) F etc. etc. 

Textus Receptus 
Chart 1 Stemma illustrating Streeter's theory of local texts (from B. H. 
Streeter, The Four Gospels, p. 26). 

23. B. IT. Streeter, The Four Gospels, A Study of Origins (London, 
1924), p. 106. 




Italy and Gaul 


Primary authority 


Syr. S. 

8565 Mk 


yMY, Mr 

Secondary do 

« L Sah. Boh. 

Syr. C. 

1 &c. 13 &c. 28 700 
(W Mk ) Old Georgian 

b a 

(W Mk ) e 

Tertiary do 

C, 33, W Lk - |,n 

AMk mMk 

Frags.: T Lk J n Z Ml H Lk 

Syr. Pesh. 

1424 &c. 544 
N-2-O <P 


fp If*'- ir C m -S n 
Frag.: n (cf. a) 

c Mk, l.k 


579 MkLkjn 892 1241 


Syr. Hcl. 
Syr. Flier. 

U A 1071 1604 
Old Arm. 




Origen A.D. 230 
Cyril Alex. 430 

Origen A.D. 240 
Rusebius 325 

Tatian 170 
Irenaeus 185 

Cyprian 250 

l&c, = 1—22— 118— 131— 209— 872 Mt — 1278— 1582— 3193. H &c. = 13— 69— 124— 230— 346— 543— 788— 826— 328— 983— 1689 — 1709. WAM~ = 28 MSS., 

including M, cited by Soden as 1*. Byzantine Te«: S V Q; F ¥ G H-, (A, K /I, Y); <D; («*"). 

Afored /-'rags. P Q R ,J< /V.fl — 1 &c. = Jam. 1 = Sod. P; H &c. = jW 13 = Sod. I', Sod. I" misleaclingly includes D with e, 28, W, 565, 700. 

Chart 2 Chart of witnesses and the local texis (from B. H. Streeter, The Four Gospels, p. 108). 

218 The Text of the New Testament 

local text support a reading, very little is gained by citing additional 
evidence of manuscripts which normally support the same local text. 21 
3. Though on minor points of reading, absolute certainty may often be 
unattainable, a text of the Gospels can be reached, the freedom of 
which from serious modification or interpolation is guaranteed by the 
concurrence of different lines of ancient and independent evidence. 25 

Today, there are few critics who follow Streeter's method in all 
of its particulars. Most recognize that the Caesarean text, for example, 
cannot be firmly established (see p. 278) and that, in fact, apart from 
the text of Alexandria, it is difficult to find clearly demarcated textual 
families associated with the various major sees of early Christendom. 
At the same time, it should be recognized that Streeter performed a 
valuable service in showing how classical methods of manuscript 
classification can be used to establish the history of textual transmis- 
sion, for the history of the text plays a major role in establishing the 
earliest form of the text. 

IV. Alternative Methods of 
Textual Criticism 

1 . The Majority Text 

The end of the twentieth century saw a resurgence of interest in 
the Byzantine text type among those who believe that the original 
text is best preserved in the vast majority of witnesses produced 
in the Middle Ages. 26 This preference for the "majority text" can be 
found among a small but vocal group of critics who, like Dean J. W. 
Burgon in a previous century (see pp. 181-2), rejected Hort's view 
that the Syrian or Byzantine text is a later recension characterized by 
inferior, secondary readings and believed that God has preserved the 
essential purity of the type of text that lies behind the King James 

24. Streeter, op. cit., p. 78. 

25. Ibid., p. 148. 

26. See the summary and critique of Daniel B. Wallace, "The Majority 
Text Theory: History, Methods, and Critique" in The Text of the New Testa- 
ment in Contemporary Research, ed. by Hhrman and Holmes, pp. 297-320. 

Modern Methods of Textual Criticism 219 

Version (Hills, 27 e.g., argued for the genuineness of even the Comma 
Johanneum of 1 John 5.7-8). 

Another group of scholars who also prefer the Byzantine text but 
recognize that the Textus Receptus is only one form of several com- 
peting forms of the Byzantine text present what appears to be a more 
sophisticated approach. 28 In their edition of The New Testament in 
the Original Greek According to the Byzantine/Majority Textform, for 
example, Maurice Robinson and William Pierpont maintain that "from 
a transmissional standpoint, a single Textform would be expected to 
predominate among the vast majority of manuscripts in the absence 
of radical and well-documented upheavals in the manuscript tradi- 
tion." 29 On this ground, they insist that it is always most probable 
that the majority of witnesses— which "reflect a high degree of tex- 
tual uniformity" — will preserve the original form of the text. 30 

27. Besides Hills' publications (see p. 182, n. 23), essentially the same 
point of view was presented by David Otis Fuller in True or False? (Grand 
Rapids, MI, 1973) and in a series of articles written by various authors in the 
Journal of the Evangelical Theological Society, xxi (1978). For a balanced re- 
buttal, see D. A. Carson, The King James Version Debate: A Plea for Realism 
(Grand Rapids, MI, 1979). Two more recent apologiae for the majority text 
and the King James Version are the truculent comments of Gordon H. Clark 
in Logical Criticisms of Textual Criticism (Jefferson, MD, 1986), and the 
one-sided symposium entitled The Majority Text: Essays and Reviews in the 
Continuing Debate, ed. by Theodore P. Letis (Fort Wayne, IN, 1987). Letis 
has a Ph.D. from the University of Edinburgh in ecclesiastical history; much 
of his dissertation has been incorporated into his volume entitled The Eccle- 
siastical Text: Text Criticism, Biblical Authority and the Popular Mind, of 
which the 2nd ed. has appeared (Philadelphia, 2000). 

28. E.g., Jakob van Bruggen, The Ancient Text of the New Testament 
(Winnipeg, 1976); Wilbur N. Pickering, The Identity of the New Testament 
Text (Nashville, 1977; 3rd ed., Eugene, OR, 2003); Zane C. Hodges and 
Arthur L. Farstad, eds., The Greek New Testament According to the Majority 
Text (Nashville, 1982; rev. ed. 1985). 

29. Preface to the 2nd ed. It is important to note that, as with other 
supporters of the majority text, Pierpont and Robinson acknowledge their 
theological assumption that God must have preserved the text for his 
church in a relatively pristine form and that the majority of witnesses are 
therefore more likely to be original. 

30. The authors object to the use of an eclectic method to establish 
the original text, on the ground that the text that is produced (e.g., in 

the United Bible Societies [UBS] Greek New Testament) is one that cannot 

220 The Text of the New Testament 

A similar position is promoted by Zane Hodges and Arthur 
Farstad, the editors of the The Greek New Testament According to 
the Majority Text, who argue their case on the basis of a similar theo- 
retical presupposition: "In any tradition where there are not major 
disruptions in the transmissional history, the individual reading 
which has the earliest beginning is the one most likely to survive in 
a majority of documents." 31 

History tells us, however, that, so far from there being no major 
disruptions in the process of transmission, during the pre-Constantinian 
persecutions New Testament manuscripts were sought out and burned 
by imperial order. Fortunately, some collections of Christian books 
escaped Diocletian's systematic program of destruction; one was the 
large collection at Caesarea, a library utilized by Origen, Eusebius, 
and even Jerome. But this too was destroyed later by Muslims in the 
year 638. The further spread of Islam in the seventh century meant 
that Christians in three of the five ancient patriarchates (Alexandria, 
Jerusalem, and Antioch) came under the sway of Muslims and the 
Christian populations of North Africa, Egypt, Palestine, Syria, and 
Mesopotamia were greatly reduced, with corresponding effects upon 
the transmission of the Scriptures in those areas. Furthermore, 
though at one time Greek had been the lingua franca of the Roman 
Empire, by the sixth century it was scarcely understood beyond the 
borders of the Byzantine Empire. So, while it is true that about 90% 
of the extant New Testament manuscripts possess a text of Byzantine 
character, it is also true that these were written after the restriction of 
Greek to basically the confines of Byzantium. 

There is, furthermore, empirical evidence that major disruptions 
have taken place in the transmission of the text of the New Testament. 
As was pointed out earlier (p. 127), more than once we find Church 

be found in any single manuscript of the New Testament but instead 
represents a modern construct cobbled together by scholars (preface to 
the 2nd ed.). As impressive as this criticism sounds, it must be pointed 
out that the same charge may be leveled against their text as well, in that 
they have established it eclectically by choosing from among the various 
readings attested in different Byzantine manuscripts, using criteria of evalua- 
tion that do not differ significantly from those used by the editors of the UBS 
text. As they themselves state in enunciating their principles: "at all times [in 
selecting the text to print] pertinent transmissional, transcriptional, external, 
and internal factors are considered as component elements of weight." 
31. Hodges and Farstad, op. cit., pp. xi-xii. 

Modern Methods of Textual Criticism 221 

fathers making reference to variant readings that were once widely 
known but are today found in only a few witnesses or even in no ex- 
tant manuscript. 32 Such a situation rules out any attempt to settle 
questions of text by adding up the numbers of witnesses in support 
of one or another variant reading. The upshot of all this is that, though 
one can be grateful to Pierpont and Robinson, on the one hand, and 
to Hodges and Farstad, on the other, for preparing editions of the 
majority text, which represent more precisely than does the Textus 
Receptus what was the prevailing form of the Greek text in the 
Byzantine period, 33 their editions are far from reproducing the origi- 
nal text of the New Testament. 34 

A mediating contribution to the ongoing debate about the 
Byzantine text is presented in the book The Byzantine Text-Type 
and New Testament Textual Criticism by the late Harry A. Sturz. 35 
The author argues that the Byzantine text is neither the original nor 
an entirely secondary text; it is an early, independent text that de- 
serves as much attention and respect as the Alexandrian and Western 
types. About a third of the book presents extensive lists, tables, and 
charts that allow the reader to assess the evidence of about 150 pas- 
sages throughout the New Testament where typical Byzantine read- 
ings are supported also by one or more of the early papyri. On the 
basis of such instances, Sturz concluded that the Byzantine text 
derives from at least the second century and represents a stream of 
tradition independent of other early traditions. 

32. For examples of variant readings that once were in the minority 
but today are dominant, and vice versa, see G. Zuntz, The Text of the 
Epistles, a Disquisition upon the Corpus Paulinum (London, 1953), p. 84. 

33- Daniel B. Wallace makes a case for using this edition in place of 
the Textus Receptus as a collating base in his article "The Majority Text: A 
New Collating Base?" New Testament Studies, xxxv (1989), pp. 609-18. 

34. For other critiques of the majority text as edited by Hodges and 
Farstad, see Michael W. Holmes, "The 'Majority Text Debate': New Form of 
an Old Issue," Tbemelios, viii, 2 (Jan. 1983), pp. 13-19; Harold P. Scanlin, 
"The Majority Text Debate; Recent Developments," The Bible Translator, 
xxxvi (1985), pp. 136-40; Kurt Aland, "The Text of the Church?" Trinity 
Journal, viii (1987), pp. 131 — 44; Daniel B. Wallace, "Some Second Thoughts 
on the Majority Text," Bibliotheca Sacra, cxlvi (1989), pp- 270-90. 

35. (Nashville, 1984). A useful feature of Sturz's book is the extensive 
bibliography of more than 800 books and articles (almost all of them 
restricted, however, to works in English), along with an index to the 
bibliography with respect to subject matter and scriptural passage. 

222 The Text of the New Testament 

Unfortunately, few of the 150 variant readings that Sturz lists are 
distinctively Byzantine; most of them have significant non-Byzantine 
witnesses supporting them as well. Moreover, one must also ask 
whether the evidence of this or that Byzantine reading among the 
early papyri demonstrates the existence of the Byzantine text type. A 
text type involves a particular constellation of readings in a charac- 
teristic pattern, and the fact is that not one of the papyri collated by 
Sturz can be characterized as Byzantine in the text that it presents. 
Other serious questions remain, most especially why patristic writers 
prior to Basil the Great and Chrysostom show no acquaintance with 
the Byzantine text. In short, one is led to conclude that Sturz failed 
to prove that the Byzantine text type is older than the fourth century. 

2. Thoroughgoing Eclecticism 

Dissatisfied with the results achieved by weighing the external 
evidence for variant readings in terms of support from individual 
manuscripts or families of manuscripts and local texts, several scholars 
have directed primary attention to the individual variants themselves 
in an effort to find which will account best for the rise of the others. 36 
This process has been given various names. Sometimes it has been 
referred to as "rational criticism." 37 The use of the adjective rational 
in this connection is not intended to suggest that all other methods 
of criticism are irrational, but that the critic is concerned primarily 
with finding plausible reasons based on internal considerations to 
justify in each case the choice of one reading as original and the 
others as secondary. More often, the method has been called "eclec- 
ticism" 38 because in its application the textual critic pays less atten- 
tion to questions of date and families of manuscripts than to internal 

36. For a recent overview by one of the principal proponents of the 
method, see J. Keith Elliott, "Thoroughgoing Eclecticism in New Testament 
Textual Criticism," in The Text of the New Testament in Contemporary 
Research, ed. by Ehrman and Holmes, pp. 321-35. 

37. M.-J. Lagrange used this term in the title of his monumental 
volume La critique textuelle; ii, La critique rationnelle (Paris, 1935), though 
in practice he frequently paid more attention to external evidence than one 
would have thought likely. 

38. E.g. G. D. Kilpatrick, journal of Theological Studies, xliv (1943), 
p. 36 ("rigorous eclecticism"), and Journal oj 'Theological Studies, xlv 
(1944), p. 65 ("impartial eclecticism"). For an example of the application 
of such eclecticism, see Kilpatrick, "An Eclectic Study of the Text of 

Modern Methods of Textual Criticism 223 

or contextual considerations. Consequently, the editor of a text follows 
now one and now another set of witnesses in accordance with what 
is deemed to be the author's style or transcriptional hazards. The 
ultimate disregard for the external evidence of variants is to be seen 
in several of Paul Mazon's editions of the Greek classics, where the 
apparatus supplies only the variants without mentioning the identity 
of the witnesses that support them. 39 

In recent times, the method has been more frequently called 
"thoroughgoing eclecticism," to differentiate it from the most com- 
mon method of textual criticism ("reasoned eclecticism"), which 
chooses the best reading by giving weight to both external and in- 
ternal evidence. Thoroughgoing eclecticism, however, gives almost 
exclusive consideration to the style of the author and the demands 
of the context. Such, for example, was the method that Bernhard 
Weiss followed in preparing his edition of the Greek Testament (see 
pp. 183—4). Likewise, C. H. Turner, having made a thorough study of 
Marcan usage, reconstructed the Greek text of the first chapter of 
Mark in accordance with stylistic considerations. 40 One of the con- 
clusions to which Turner's investigations led him was the need to 
show greater respect for Western readings, even though they may be 
supported by only a few witnesses (e.g., by D and one of the three 
leading Old Latin manuscripts, k, e, or a). 41 

Acts," in Principles and Practice of New Testament Textual Criticism: 
Collected Essays of G. D. Kilpatrick, ed. by J. K. Elliott (Leuven, 1990), 
pp. 358-69. 

39. See A. Dain's comments on "Olympian indifference" to what is 
regarded as "useless erudition" (Les Manuscrits [Paris, 19491, p. l6l; rev. 
ed. [1964], p. 175). 

40. C. H. Turner, "A Textual Commentary on Mark 1," Journal of 
Theological Studies, xxviii (1926-7), pp. 145-58; idem, "Marcan Usage: 
Notes, Critical and Exegetical, on the Second Gospel," Journal of 
Theological Studies, xxv (1923-4), pp. 377-86; xxvi (1924-5), pp. 12-20, 
145-56, 225-40, 337-46; xxvii (1925-6), pp. 58-62; xxviii (1926-7), 

pp. 9-30, 349-62. 

41. See idem, "Western Readings in the Second Half of St. Mark's 
Gospel," Journal oj "Theological Studies, xxix (1927-8), pp. 1-16; idem, 
"The Textual Criticism of the New Testament," in A New Commentary on 
Holy Scripture . . . , ed. by Charles Gore et al. (London, 1928), pp. 718-29. 
For adumbrations of this point of view, compare Turner's earlier article, 
"New Testament, Text of," in The Illustrated Bible Dictionary, ed. by 
William C. Piercy (New York, 1908), pp. 585-96. 

224 The Text of the New Testament 

So too, several scholars have examined the text of certain other 
New Testament books in the light of thoroughgoing eclecticism. For 
the Pauline Epistles, especially for 1 Corinthians and Hebrews, the 
lively and learned Schweich Lectures delivered in 1946 by Gimther 
Zuntz have much to teach the student regarding text-critical 
method. 42 As regards the Book of Revelation, one of the most valu- 
able sections of Josef Schmid's extensive monograph on the history 
of the Greek text of the Apocalypse is his chapter dealing with the 
bearing of the linguistic usage of the book on the textual evaluation 
of variant readings. 43 

Turner's reliance on stylistic criteria in making textual decisions 
was revived by George D. Kilpatrick of Queen's College, Oxford, 
and his student, J. Keith Elliott, of Leeds. Portions of Kilpatrick's re- 
search into questions of lexical and grammatical usage of authors of 
the New Testament were published in several periodicals, 44 and his 
text-critical conclusions found expression in fascicles of A Greek- 
English Diglotforthe Use of Translators, issued for private circulation 
by the British and Foreign Bible Society (Mark, 1958; Matthew, 1959; 
John, I960; The General Letters, 1961; Luke, 1962; The Pastoral Letters 
and Hebrews, 1963; and Romans and 1 and 2 Corinthians, 1964). 
Of two or more variant readings, Kilpatrick usually preferred that 
which accords with what is deemed to be the author's style, irre- 
spective of the date and nature of the external evidence supporting 
the reading. In matters on which no firm decision can be made con- 
cerning the author's style, he often appealed to the criterion of Atti- 
cism, which became one of the dominant tendencies in literary 
circles during the second Christian century. He argued that scribes in 
the second century introduced many Atticisms into the text of the 

42. G. Zuntz, The Text of 'the Epistles: A Disquisition upon the Corpus 
Paulinum (London, 1953). 

43. Josef Schmid, Studien zur Geschichte des griechischen Apokalypse- 
Textes; ii, Die alten St&mme (Munich, 1955), pp. 173—251 ■ 

44. G. D. Kilpatrick, "Some Notes on Marcan Usage," Bible Translator, 
vii (1956), pp. 2-9, 51-56, 146; idem, "Some Notes on Johannine Usage," 
Bible Translator, xi (I960), pp. 173-7; "AiaXeyeaBai and dicdoyi^eadai 

in the New Testament," Journal of Theological Studies, n.s., xi (I960), 
pp. 338-40. These and others of Kilpatrick's articles can now be found in 
the collection The Principles and Practice of New Testament Textual 
Criticism, ed. by Elliott. 

Modern Methods of Textual Criticism 225 

New Testament/ 5 Of two readings, therefore, of which one con- 
forms to Attic canons and the other does not, he was inclined to 
accept the non-Attic reading, even though no early manuscript 
evidence may support it. In order to justify his general disregard for 
the age and quality of external evidence, Kilpatrick declared that by 
about A.D. 200 the great majority of the deliberate changes had been 
introduced into the textual stream and that thereafter scribes trans- 
mitted the several forms of text with great fidelity. Thus, though a 
variant reading may happen to be preserved only in a late minuscule 
manuscript, if it is in harmony with what is taken to be the author's 
style or reflects a non-Attic tendency, Kilpatrick is disposed to regard 
it as original. 46 

The extent to which Kilpatrick is prepared to go in adopting read- 
ings that have the most meager external support, if he is convinced 
that internal considerations require it, may be illustrated by the 
following readings in his Greek- English Diglot: 

Matt. 20.30 exgat,ov is supported by 118, 209, Syr c - pal . 

Matt. 22.1 omit ebievwith E, Syr p . 

Matt. 22.7 axovaag de 6 fiaotXevg ixelvog is supported by 33- 

Mark 5.11 rd ogrj is supported by 372, 485, Syr 5 . 

Mark 9.17 dnoxQiOeig avrq> is supported by C. 

Mark 14.6 eig ifie is supported by 517, 579, /251, Syr*' p , Eth. 

Mark 14.31 iXdXet /udXXov is supported by 574 and k. 

Luke 9-51 ioriJQi&v is supported by 1241. 

John 19.35 dXrjOr/g is supported bys, 124, Chr. 

Jam. 2.18 egycov rr/v tiiouv ptov is supported by Syr 11 . 

1 Pet. 2.11 ojiExeoOai Vfidg is supported by Vulg, Cyp. 

2 John, vs. 8 JiXrjgtjg is supported by L, Dam. 

By way of summary, it is obvious that there is much to commend 
the practice of a judicious eclecticism in text criticism, for no one 

45. See Kilpatrick's article, "Atticism and the Text of the Greek New 
Testament," in Neutestamentliche Aufsatze: Festschrift fur Prof . Josef 
Schmid (Regensburg, 1963), pp. 125-37; also his comments in Gottingische 
gelehrte Anzeigen, ccxv (1963), pp. 14 ff. 

46. Similar principles are enunciated and followed by J. K. Elliott; 
see especially his essay on "thoroughgoing eclecticism" in Text of the 
New Testament in Contemporary Research, ed. by Ehrman and Holmes, 
pp. 321-35. 

226 The Text oe the New Testament 

manuscript and no one family preserves the original text in its en- 
tirety. Certainly, the critic must always take into account what is or is 
not the usage of a given author, for, as A. E. Housman once expressed 
it in a pithy epigram: "The indulgence of love for one manuscript and 
dislike for another inevitably begets indifference to the author him- 
self." 47 At the same time, however, the weaknesses inherent in a thor- 
oughgoing eclecticism should not be overlooked. Statistics regarding 
an author's usage are sometimes derived from concordances that are 
based on editions of the Greek Testament that contain, in some 
passages, quite indefensible readings. Furthermore, even if evidence 
regarding an author's usage has been sifted critically, its significance 
for the passage at hand must be weighed in the light of two possibil- 
ities: (1) an author may on occasion vary usage and (2) a scribe who 
was aware of the author's prevailing usage may have altered a read- 
ing in order to bring it into harmony with that usage. 

One may conclude, therefore, that although consideration of the 
literary usage of a New Testament author will be of considerable 
value to the textual critic, it must not be made the primary criterion 
in the evaluation of variant readings to the virtual neglect of external 
evidence. Furthermore, to apply rigorously considerations based on 
the Attic revival during the early Christian centuries is to be in danger 
of disregarding the operation of other literary and stylistic tendencies 
that also influenced the Koine during the same period, some of which 
were deliberately anti-Attic. 48 

V. Conjectural Emendation 

The classical method of textual criticism regularly involves, as was 
mentioned at the beginning of this chapter, the exercise of conjectural 
emendation. If the only reading, or each of several variant readings, 
that the documents supply is impossible or incomprehensible, the 

47. In Housman's edition of Lucan's Bellum civile (Oxford, 1926), p. vi. 

48. See, e.g., the references under "Antiatticismus" in Wilhelm 
Schmid's Der Atticismus in seinen Hauptvertretern (Stuttgart, 1887-96) and 
cf. Radermacher's warning against undue reliance upon statistical consider- 
ations in judging the literary style of an author in a period when quite 
opposite influences cut across one another; Ludwig Radermacher, Koine 
(Sitzungsbericbte der Akademie der Wissemcbaften in Wien, Phil. hist. Kl., 
ccxxiv, 5; Abhandlung, 1947), pp. 61 f. 

Modern Methods of Textual Criticism 227 

editor's only remaining resource is to conjecture what the original 
reading must have been. 

A typical emendation involves the removal of an anomaly. It must 
not be overlooked, however, that though some anomalies are the 
result of corruption in the transmission of the text, others may have 
been either intended or tolerated by the author himself. 49 Before re- 
sorting to conjectural emendation, therefore, the critic must be so 
thoroughly acquainted with the style and thought of the author that a 
certain anomaly must be judged to be foreign to the author's intention. 

This aspect of criticism has at times been carried to absurd 
extremes. In his later work, Richard Bentley, for example, largely 
disregarded the evidence of manuscripts in determining the correct 
readings and depended chiefly on his own instinct as to what an 
author must have written. He justified such a procedure in the 
magisterial phrase "nobis et ratio et res ipsa centum codicibus potiores 
sunt," which may be rendered "for me both reason and the subject 
matter are worth more than a hundred manuscripts." In following this 
bold principle, he did much that was rash and indefensible as well as 
much that was brilliant and convincing. The reductio ad absurdum of 
such a subjective method is found in Bentley's edition of Milton's Par- 
adise Lost, in which he offers more than 800 emendations, restoring 
what in his opinion Milton must have really said (or meant to say) 
while dictating the poem to his daughters. 50 

Before a conjecture can be regarded as even probable, it must 
satisfy the two primary tests that are customarily applied in evaluat- 
ing variant readings in manuscripts: (1) it must be intrinsically suit- 
able and (2) it must account for the corrupt reading or readings 
in the transmitted text. There is, however, an important difference 
between the method of applying these tests to a conjectural emen- 
dation and that of applying them to variants in manuscripts. We 
accept the variant that best satisfies the tests, but we require of a 
successful conjecture that it shall satisfy them absolutely well. The 
conjecture does not rise from a certain level of probability (a "happy 
guess") to the level of certainty, or approximate certainty, unless its 

49. For a discussion of the paradoxical possibility of a textual critic's 
"improving" on the original, see G. Zuntz's article on 1 Cor. 6.5, "The Critic 
Correcting the Author," Philologus, xcix (1955), pp. 295-303. 

50. See James Henry Monk, The Life of Richard Bentley, D.D., 2nd ed., 
ii (London, 1833), pp. 309-23; Richard C. Jebb, Bentley (London, 1889), 
pp. 180-91. 

228 The Text of the New Testament 

fitness is exact and perfect. The only criterion of a successful con- 
jecture is that it shall approve itself as inevitable. Lacking inevitabil- 
ity, it remains doubtful. 

An example from English literature will illustrate the wide differ- 
ences of merit among proposed conjectures. 51 Since the early printers 
in England were often foreigners, 'who made quite as bad mistakes as 
their predecessors the scribes, the text of Shakespeare contains al- 
most as many problem passages as that of Aeschylus. In the folio edi- 
tions of Henry V, act 2, scene 3, the hostess says of the dying Falstaff 
"his nose was as sharp as a pen and a table of Green Fields." The 
words "a table of Green Fields," which appear with trifling variations 
of spelling in the folio editions but are omitted in the quarto editions, 
have been the subject of numerous conjectural emendations. Pope 
suggested (perhaps ironically) that this was a stage direction to bring 
in one of Greenfield's tables, Greenfield being supposed to be the 
furniture dealer who supplied props for Shakespeare's theater. Collier 
proposed "on a table of green frieze," and another critic suggested "or 
as stubble on shorn fields." The conjecture that today is adopted by 
editors is "and a' babbled of green fields," being a modification by 
Theobald of a happy proposal made by an anonymous annotator 
who corrected "a table" to "a' talked." 52 

The fault most often committed in the use of conjectural emen- 
dation has been to use it prematurely. Corruptions in the Greek and 
Latin classics (including the New Testament) have frequently been 
assumed without adequate reason — as though, indeed, for the mere 
sake of showing off one's cleverness in proposing an alternative 

51. This example is taken nearly verbatim from James Gow's 
Companion to School Classics, 2nd ed. (London, 1889), pp. 65 f. 

52. Several passages in Shakespeare are corrupt beyond the ingenuity 
of paleographer and textual critic to propose a cure. Apart from lucky 
coincidence, what lay behind the hodgepodge of nonsense set by the 
compositor of the first quarto of King Lear'm act 3, scene 4, pp. 118 ff., is 
probably unattainable: "swithald footed thrice the old a nellthu night more 
and her nine fold bid her, O light and her troth plight and arint thee, with 
arint thee." On the special problems involved in the textual criticism of 
Shakespeare's works, see Madeleine Doran, "An Evaluation of Evidence in 
Shakespearean Textual Criticism," in English Institute Annual, 1941 (New 
York, 1942), pp. 95-114; F. P. Wilson, "Shakespeare and the 'New 
Bibliography,'" in The Bibliographical Society, 1892-1942, Studies in 
Retrospect (London, 1945), pp. 133-4. 

Modern Methods of Textual Criticism 229 

reading. This "itch for emending" {pruritus emendandt) has resulted 
in the accumulation of literally thousands of proposed alterations 
of passages in the New Testament. Those which William Bowyer 
assembled in the eighteenth century (see p. 162) were greatly aug- 
mented in the latter part of the nineteenth century by a spate of arti- 
cles and books published in Holland by W. C. van Manen, W. H. van 
der Sande Bakhuyzen, D. Hailing, S. S. de Koe, H. Franssen, J. M. S. 
Baljon, J. H. A. Michelsen, J. Cramer, and others. 53 

In their edition of the Greek New Testament, Westcott and Hort 
marked with obeli about 60 passages that they (or one of them) 
suspected involve a "primitive error," that is, an error older than 
the extant witnesses, for the removal of which one is confined to 
conjectural emendation. 54 According to Schmiedel, 55 the editions of 

53- For the titles of these works, see Eberhard Nestle's bibliographical 
list in Urtext unci Obersetzungen der Bibel (Leipzig, 1897), pp. 55-6 (a repr. 
of "Bibeltext und Bibeliibersetzungen" in Herzog-Hauck's Realencyklopddie 
fur protestantische Theologie und Kirche, 3te Aufl.). In the twentieth 
century, the Dutch philologist I. I. Hartman maintained that most errors 
in the transmission of the classics are nonmechanical and therefore inex- 
plicable and that the editor of a text is permitted to abandon manuscript 
evidence whenever in his or her opinion this is demanded by the content; 
see I. I. Hartman, "Ars critica, quid sibi habeat propositum et qua utatur 
ratione," Mnemosyne, n.s., xlviii (1920), pp. 227-38. For a rebuttal of 
Hartman's argument, see A. Damste, "De arte critica," Mnemosyne, 
pp. 424-33; and for several discussions of the proper limitations of 
conjectural emendation, see the appendix added to the second edition of 
Fasquali's Storia della tradizione e critica del testo, entitled "Congettura 
e probability diplomatics, " pp. 481-6; Paul van den Ven, "Erreurs de 
methode dans la correction conjecturale des textes byzantins," Byzantion, 
xxiv (1954), pp. 19-45; and the sane remarks by Ludwig Bieler in his 
stimulating essay "The Grammarian's Craft," Folia: Studies in the 
Christian Perpetuation of the Classics, x (1956), pp. 3-42, especially pp. 26 f. 

54. The following are the passages where Westcott and Hort sus- 
pected the presence of a primitive error: Matt. 21.28 ff.; 28.7; Mark 4.28; 
Luke 11.35; John 4.1, 6.4, 8.9; Acts 4.25, 7.46, 12.25, 13.32, 43, 16.12, 
19.40, 20.28, 25.13; Rom. 1.32, 4.12, 5.6, 7.2, 13-3, 15.32; 1 Cor. 12.2; 2 Cor. 
3.3, 17, 7.8, 12.7; Gal. 5.1; Col. 2.2, 18, 23; 2 Thess. 1.10; 1 Tim. 4.3, 6.7; 2 
Tim. 1.13; Philem. 9; Heb. 4.2, 10.1, 11.4, 37, 12.11, 13.21; 1 Pet. 1.7, 3.21; 
2 Pet. 3.10, 12; 1 John 5.10; Jude 1, 5, 22 f.; Rev. 1.20, 2.12, 13, 3.1, 7, 14, 
9.10, 11.3, 13.10, 15, 16; 18.12, 19.13. 

55. Paul W. Schmiedel in Festgabe Adolf Kaegi von Schulern und 
Freunden dargebracht zum 30. September 1919 (Frauenfeld, 1919), p. 179. 

230 The Text of the New Testament 

Tregelles, Tischendorf, and Weiss contain only one conjecture each: 
Tregelles at 1 Pet. 3.7 margin; Tischendorf at Heb. 11.37; and Weiss 
at Rev. 18.14. The apparatus of the 24th edition of Nestle's Greek 
New Testament includes from various sources about 200 conjectures, 
90 of which are identified by the name of the scholar who first sug- 
gested them; in the 27th edition, there are 129 conjectures noted, all 
but one of them identified by the scholar who first proposed them. 56 
One must admit the theoretical legitimacy of applying to the 
New Testament a process that has so often been found essential in 
the restoration of the right text in classical authors. But the amount 
of evidence for the text of the New Testament, whether derived from 
manuscripts, early versions, or patristic quotations, is so much 
greater than that available for any ancient classical author that the 
necessity of resorting to emendation is reduced to the smallest 
dimensions. ,7 Thus, whereas several scholars during the eighteenth 
and nineteenth centuries amused themselves by proposing thou- 
sands of conjectural emendations for various passages of the New 
Testament, the tendency in recent days has been to exercise much 

56. Information on the Nestle-Aland 27th ed. has been provided by 
Jan Krans of Amsterdam, whose recently competed Ph.D. dissertation deals 
with conjectural emendations in the New Testment. 

57. Of all the emendations proposed for the New Testament, 
perhaps the one that has enjoyed the widest favor is the suggestion that 
the name "Enoch" has fallen out of the text of 1 Pet. 3- 19, ''[the spirit] in 
which he [Christ or Enoch?] went and preached to the spirits in prison." 
From 1772, when Bowyer included this emendation in the second edition 
of his Critical Conjectures, down to the present, a large number of 
scholars have given it their approval, including James Moffatt and Edgar 
J. Goodspeed, both of whom introduced it into the text of their 
translations of the New Testament. (For the history of this conjecture and 
a list of scholars who have adopted it, see B. M. Metzger in Journal of 
Religion, xxxii [1952], pp. 256 f.) What shall be said of this conjecture? 
Admittedly, the precise meaning of 1 Pet. 3.19 in the generally received 
text is difficult to ascertain, and a great number of interpretations have 
been proposed; also, the proposed emendation is attractive paleographi- 
cally (ev q> xai and 'Eva>x in uncial script are remarkably similar: 
ENQKAI and ENQX). Nevertheless, since the introduction of a new 
subject ("Enoch") into verse 19 disturbs an otherwise smooth context and 
breaks the continuity of the argument, the emendation cannot be 
accepted — for an emendation that introduces fresh difficulties stands 

.Modern Methods of Textual Criticism 231 

more caution in proposing and adopting such corrections.' 8 In fact, 
of all the conjectures attributed to various scholars included in the 
apparatus of the Nestle-Aland Greek New Testament, only a handful 
are reflected in text or marginal rendering of a broad range of 
English, French, and German translations and revisions of the New 
Testament. 59 

VI. Methods of Determining Family 
Relationships Among Manuscripts 60 

Since 1725, when Johann Albrecht Bengel first proposed that the 
surviving manuscripts of the New Testament could be classified 
in "companies, families, tribes, and nations" (see p. 159), scholars 
have recognized the value of determining the family relationships 
among our surviving manuscripts. For one thing, it is completely im- 
practicable to consult every one of the thousands of textual wit- 
nesses when trying to determine the oldest form of the text where 
there are variant readings. Since the manuscripts are related to one 
another in textual groups, it is possible simply to consult the leading 
witnesses of a group in order to determine the family reading. More- 
over, once manuscripts are organized according to their textual 
affinities, the groups themselves can be evaluated with respect to the 
kinds of reading they have in common where manuscripts of other 
groups attest variation. On the basis of these evaluations, one group 

58. John Strugnell's "A Plea for Conjectural Emendation in the 
New Testament," Catholic Biblical Quarterly, xxxv (1974), pp. 543-58, 
received a strong and vigorous rebuttal from G. D. Kilpatrick, "Conjectural 
F.mendation in the New Testament," in New Testament Textual Criticism, Its 
Significance for Exegesis, ed. by E.J. Epp and G. D. Fee (Oxford, 1981), 
pp. 349-60. See also David Alan Black's judicious analysis "Conjectural 
Emendations in the Gospel of Matthew," Novum Testamentum, xxxi 
(1989), pp. 1-15. 

59. See Erroll F. Rhodes, "Conjectural Emendations in Modern 
Translations," in New Testament Textual Criticism, ed. by Epp and Fee, 
pp. 361-80. See also P. W. Comfort, Early Manuscripts and Modern 
Translations of the New Testament (Wheaton, IL, 1990). 

60. See B. D. Ehrman, "Methodological Developments in the Analysis 
and Classification of New Testament Documentary Evidence," Novum 
Testamentum, xxix (1987), pp. 22-45. 

232 The Text of the New Testament 

may be seen to approximate the original text more closely than the 
others. Thus, for example, if one group typically attests readings that 
are harmonized with other passages or that are conflated from read- 
ings of other parts of the textual tradition, this group would be less 
valuable, as a rule, for determining the original text. Conversely, a 
group that normally does not attest secondary harmonizations or 
conflations is more likely to preserve the original text. Thus, in the 
traditional approach of classification, once manuscript categories 
have been established on the basis of textual consanguinity, they can 
be evaluated qua groups and so be used in making textual decisions. 

Another way critics use textual groupings is by considering groups 
not in terms of their individual superiority but in their patterns of 
combination with one another in support of a given reading. Thus, for 
example, a reading found only in primary Alexandrian and Western 
texts is commonly judged original, given the early dates and apparent 
independence of these two textual groups. 

Despite the widely recognized value of determining textual 
groupings, it has been only in relatively recent times that scholars 
have developed methods of classifying New Testament witnesses 
according to scientific principles. In the early days of the discipline, 
scholars typically classified witnesses on an ad hoc basis, by compar- 
ing several key readings of manuscripts with individual witnesses that 
were thought to be representative of different families or types of text. 
More often, the manuscript was collated against the Textus Receptus; 
the variant readings that were uncovered were then analyzed in terms 
of agreement with other manuscripts whose readings were reported 
in various apparatus critici. This procedure made sense to scholars 
who understood the Textus Receptus to represent the original text of 
the New Testament, for then variations from it would be "agreements 
in error" and would indicate the genealogical relations of the wit- 
nesses that attest them. The procedure continued to make sense after 
the overthrow of the Textus Receptus in the late nineteenth century, 
for then variations from it were understood to represent older forms 
of the text and the manuscripts that attested them could be grouped 
together by considering which of these variations they contained. 

Eventually it came to be recognized, however, that this method 
of manuscript classification was inherently flawed because it did not 
consider a large part of the evidence. Thus, it is quite deceptive to 
know only that in a certain chapter a given manuscript agrees with, 
say, 15 and R ten times in differing from the Textus Receptus; for if B 
and s* should also, in that chapter, differ from the Textus Receptus in 

Modern Methods of Textual Criticism 233 

90 other instances, the Alexandrian element in the given manuscript 
would be slight indeed. 

In order to provide a more exact method of analysis of manuscript 
relationships, in 1911 E. A. Hutton proposed what he called "triple 
readings." 61 Selecting a list of readings in which the Alexandrian, 
Western, and Syrian (Byzantine) authorities are divided, he urged 
that henceforth all manuscripts be analyzed by determining the num- 
ber of agreements with each of these three textual types. 

With the multiplication of the number of identifiable textual 
groups, it became desirable to seek a higher degree of precision than 
Hutton's method permits. E. C. Colwell, with the assistance of M. M. 
Parvis, elaborated the method of multiple readings in determining 
the relationship of manuscripts. A multiple reading is denned as one 
in which the minimum support for each of at least three variant 
forms of the text is either one of the major strands of the tradition, 
the support of a previously established group (e.g., Fam. 1, Fam. 7t, 
the Ferrar group, K 1 , K', K r ), the support of one of the ancient 
versions (e.g., af, it, sy\ sy^', bo, sa), or the support of some single 
manuscript of an admittedly distinctive character (e.g., D) 62 

Colwell applied these specifications to a sample section of text 
extending from John 1.1-4.40 and found 22 instances of multiple 
readings. By tabulating the number of agreements of a newly found 
manuscript with the several witnesses at those passages that involve 
multiple readings, one will learn something of the textual complex- 
ion of that manuscript. 

The limitation of this method, at least as applied to the sample 
from John, is that test passages are relatively few in number com- 
pared to the extent of text investigated. It is conceivable that a given 
manuscript may agree consistently with a particular type of text in 
most of the 22 passages in the 152 verses covered by John 1.1-4.40 
but disagree with that text type in a significant number of readings 
in the remaining 130 verses not represented by multiple readings. 

61. Edward Ardron Hutton, An Atlas of Textual Criticism, Being an 
Attempt to Show the Mutual Relationship of the Authorities for the Text of 
the New Testament up to About 1000 A.D. (Cambridge, 1911). 

62. E. C. Colwell, "Method in Locating a Newly-Discovered Manuscript 
within the Manuscript Tradition of the Greek New Testament," in Studia 
Evangelica . . . , ed. by Kurt Aland et al. (Texte und Untersuchungen, 
lxxiii; Berlin, 1959), p. 759, repr. in his Studies in Methodology in Textual 
Criticism of the New Testament (Grand Rapids, MI, 1969) pp. 26-34. 

234 The Text of the New Testament 

Ultimately more fruitful for the task of classifying manuscripts is 
another set of suggestions made by Colwell, in a method that has 
come to be known as the "quantitative method of textual analysis." 63 
Colwell pointed out that the method of classifying New Testament 
manuscripts in terms of how far they diverge from the Textus Receptus 
is one-sided, for it does not take into account those instances in 
which manuscripts agree with one another in readings where they 
do not diverge from the Textus Receptus. Colwell maintained that 
"text types" (i.e., broad textual groups) could be determined by col- 
lating manuscripts against a base text (it could be the Textus Receptus 
or any other text) and then comparing the total number of agreements 
and disagreements of every witness with every other in variations 
judged to be genetically significant (i.e., removing from consideration 
such insignificant variations as the movable v and the alteration of 
ovrco and ovrcog). In his view, for a group of witnesses to be con- 
sidered a text type, they need to agree in approximately 70% of all 
places of genetically significant variation and to be separated from 
witnesses of other groups by a margin of around 10%. 

For the related question of how to locate a new witness within 
one of the previously established textual families, Colwell proposed 
a similar procedure: the witness is collated completely against a 
range of others (whose affinities are already known), its agreements 
and disagreements are noted, and it is confirmed as belonging to a 
textual group if it shares a high percentage of the group's readings. 

Thus, textual groups can be identified and individual manu- 
scripts can be located within these groups on the basis of their rela- 
tionship to one another, rather than in relation to an external norm 
(the Textus Receptus). The quantitative method has proved itself 
useful in study after study as scholars employed it with remarkable 
success during the second half of the twentieth century. 64 

63. See Colwell's revised and updated essays in his Studies in Method- 
ology in Textual Criticism, especially "Method in Locating a Newly- 
Discovered Manuscript," pp. 26-44, and "Method in Establishing Quantita- 
tive Relationships Between Text-Types of New Testament Manuscripts" 
(with Ernest W. Tune), pp. 56-62. 

64. Among other things, the method has been used to demonstrate 
the Western affinities of Codex Sinaiticus in the opening chapters of John 
(Gordon Fee, "Codex Sinaiticus in the Gospel of John: A Contribution to 

Modern Methods of Textual Criticism 235 

In addition to ColwelPs quantitative method, 65 three other meth- 
ods have been developed by scholars interested in classifying the 
manuscripts of the New Testament. 66 Each of these methods has a 
different purpose, so they should be seen as complementary rather 
than as in competition with one another. 

Methodology in Establishing Textual Relationships," New Testament Stud- 
ies xv [1968-91, pp. 23-44); to trace the rise and development of the 
Alexandrian text (idem, "Origen's Text of the New Testament and the 
Text of Egypt," New Testament Studies 28 [1982], pp. 248-64; idem "p 75 , 
p 66 , and Origen: The Myth of Textual Recension in Alexandria," in New 
Dimensions in New Testament Studies, ed. by Richard N. Longenecker 
and Merrill C. Tenney [Grand Rapids, MI, 1974]; Bart D. Ehrman, Didymus 
the Blind and the Text of the Gospels [Atlanta, 1986]); to discount the no- 
tion of a "pre-Caesarean text of the Gospels" (Larry Hurtado, Text-Critical 
Methodology and the Pre-Caesearean Text: Codex W in the Gospel of Mark 
[Grand Rapids, MI, 1981]); to determine the textual affinities of the surviv- 
ing witnesses of the Johannine Epistles (W. L. Richards, Ihe Classification 
of the Greek Witnesses of the fohannine Epistles [Missoula, MT, 1977]); and 
to evaluate the textual affinities of a number of patristic sources, see 
p. 239, n. 71). 

65. See further Klaus Wachtel, "Colwell Revisited: Grouping New 
Testament Manuscripts," in The New Testament in Early Christianity: 
Proceedings of the Lille Colloquium, fuly 2000, ed. by Christian-B. 
Amphoux and J. Keith Elliott (Lausanne, 2003), pp. 31-43. 

66. Another important method for classifying manuscripts was devel- 
oped by Gerd Mink and is called the "local genealogical method." The 
method is concerned with establishing the genealogical relationships of 
readings found in a manuscript by determining which form of the text 
(wherever there is variation) is antecedent to the variation. Once this deter- 
mination is made for all the variant readings of a book, the results can be 
tabulated to determine which texts (found in certain manuscripts) must 
have been derived from which others (found in other manuscripts), 
thereby making possible the creation of a genealogical tree of manuscripts 
based on their texts. See G. Mink, "Eine umfassende Genealogie der 
neutestamentlichen L'berlieferung," New Testament Studies, xxxix (1993), 
pp. 481-99. To see how the method can work in practice, see Matthew 
Spencer, Klaus Wachtel, and Christopher J. Howe, "The Greek Vorlage of 
the Syra Harclensis: A Comparative Study on Method in Exploring Textual 
Genealogy," TC: A Journal of Biblical Textual Criticism [] 
vii (2002), pars. 13-24. 

236 The Text of the New Testament 

1. The Claremont Profile Method 

The Claremont Profile Method was developed by two students 
of Colwell at Claremont, Paul McReynolds and Fredrik Wisse. 67 
While working on the International Greek New Testament Project, 
dedicated to producing an apparatus for the Gospel of Luke, 
McReynolds and Wisse were assigned the task of classifying the 
nearly 1,400 manuscripts of Luke's Gospel so as to determine which 
witnesses could most appropriately be cited in the apparatus at 
places of variation. This was obviously a gargantuan task and could 
not well be performed by means of a quantitative method requiring 
full collations and detailed comparisons of all the manuscripts. More- 
over, since the vast majority of the manuscripts were already known 
to belong to the Byzantine tradition (von Soden's k group), full col- 
lations would reveal very high percentages of agreement, showing 
the manuscripts' basic affinities but not indicating the Byzantine 
subgroups that had to be known for representative witnesses to be 

In the course of their collations, McReynolds and Wisse began to 
see that the Byzantine witnesses share readings in certain combina- 
tions, that is, that noticeable patterns of attestation began to emerge, 
patterns that can be readily displayed in graphic form. They conjec- 
tured that subgroup affiliation can be established by determining 
the profile of readings generally supported by members of a group, 

67. For an initial statement concerning the rationale and application of 
the Claremont Profile Method, see Eldon Jay Epp, "The Claremont Profile- 
Method for Grouping New Testament Minuscule Manuscripts," in Studies 
in the History and Text of the New Testament, ed. by Boyd L. Daniels and 
M.Jack Suggs (Studies and Documents 29; Salt Lake City, UT, 1967) 
pp. 27-37; the fullest statement of the history and procedures of the 
method, along with a discussion of its application to the manuscripts of 
Luke, can be found in Frederik Wisse, The Profile Method for the Classifica- 
tion and Evaluation of Manuscript Evidence as Applied to the Continuous 
Greek Text of the Gospel of Luke (Studies and Documents 44; Grand Rapids, 
MI, 1982). For a further application of the method, see W. L. Richards, The 
Classification of the Greek Manuscripts of the Johannine Epistles (Society of 
Biblical Literature Dissertations Series 35; Missoula, MT, 1977). For a critical 
evaluation, see David. C. Parker, "A Comparison Between the Text und 
Textwert and the Claremont Profile Method Analyses of Manuscripts in the 
Gospel of Luke," New Testament Studies, li (2000), pp. 108-38. 

Modern Methods of Textual Criticism 237 

that is, when two-thirds of all group members contain the readings 
in question. Once the patterns of attestation (profiles) have been 
determined for each subgroup, other documents need not be col- 
lated in toto but only in readings that signify membership in one sub- 
group or another. When a previously uncollated document generally 
agrees with the distinctive profile of an established subgroup, it can 
be classified as a group member. 

There is no doubt that the Claremont Profile Method can accel- 
erate the otherwise long and arduous process of analyzing and clas- 
sifying New Testament manuscripts. Wisse claims that with the 
method he can classify a previously unknown manuscript of Luke in 
only 30 minutes. But it should be noted that the method cannot serve 
as a substitute for a full quantitative analysis. It is best used as origi- 
nally designed: to establish in rough-and-ready fashion the sub- 
groups of the larger Byzantine family of witnesses. 

2. The Alands' Use of Teststellen 

A totally different approach to classifying New Testament wit- 
nesses was developed at the Institut fur neutestamentliche Text- 
forschung in Miinster by Kurt and Barbara Aland. This method is 
designed neither to provide a final and definitive classification of wit- 
nesses based on their textual affinities (the goal of the quantitative 
method) nor to indicate the subgroups within the Byzantine tradition 
(the goal of the Claremont Profile Method). Instead, its purpose is to 
determine which witnesses are most valuable for establishing the 
original text of the New Testament, which (others) are most valuable 
for determining the course of its transmission, and which (others) 
can safely be relegated to textual families (Western and Byzantine) 
that do not contribute significantly to either purpose. The method is 
based on collations of witnesses in a number of "test passages" 
( Teststellen) that K. Aland selected throughout the books of the New 
Testament. A witness is collated in every one of these passages to 
see if it (1) agrees with the "original" text (in the Alands' judgment), 
(2) agrees with the Byzantine text, (3) agrees with the Byzantine text 
when that text also happens to be the "original" one, (4) attests some 
distinctive reading, or (5) agrees with the D (normally designated 
Western) text. 

As a result of the collations, manuscripts are then assigned to 
one of five categories: I, witnesses that generally attest the "original" 
text; II, witnesses that often attest the original text; III, witnesses that 

238 The Text of the New Testament 

are important for understanding the history of the text; IV, witnesses 
of the D text; V, witnesses of the Byzantine text. 68 

It should be clear that these categories, except for the final two, 
are not established strictly on the basis of textual affinities. The man- 
uscripts in category III, for example, are placed there, not because 
they happen to agree with one another in a large number of their vari- 
ations, but because the Alands find their texts, which diverge from 
the Byzantine, the Western, and the "original" texts, to be important for 
establishing the history of the tradition. Thus, these five categories do 
not assist researchers interested in knowing which family groups have 
been established based on the Teststellen. Moreover, there is a curious 
kind of circularity in this approach to classification since if one of the 
purposes of grouping witnesses is to assist in establishing the "original 
text," it makes little sense to prejudge the issue by classifying wit- 
nesses precisely by how well they attest the original text! At the same 
time, the Alands' method is successful in achieving its particular aim: 
to decide which manuscripts can be safely set aside as representing 
the Byzantine tradition and which others should then be cited in the 
critical apparatus every time they indicate places of variation. 69 

3. The Comprehensive Profile Method 

Whereas the Claremont Profile Method serves to establish sub- 
groups within the Byzantine tradition and the Alands' method indi- 
cates manuscripts that should be cited in a critical apparatus, a third 
method of classification, the Comprehensive Profile Method, seeks 
to fulfill Colwell's aim of classifying manuscripts according to tex- 
tual affinities into family groups. 70 Unlike the Claremont Profile 

68. See the lists of manuscripts assigned to their respective categories 
in Kurt Aland and Barbara Aland, The Text of the New Testament: An 
Introduction to the Critical Editions and to the Theory and Practice of 
Modern Textual Criticism, trans, by Erroll F. Rhodes (Grand Rapids, ML 
1987), pp. 83-160. 

69. For an example of the Alands' method of classification used in 
conjunction with a quantitative method, so that a manuscript selected by 
Teststellen is then evaluated by full collations against a range of witnesses, 
see Spencer, Wachtel, and Howe, op. cit., pars. 4-12. 

70. See Bart D. Ehrtnan, "The Use of Group Profiles for the 
Classification of New Testament Documentary Evidence," fournal of Biblical 
Literature, cvi (1987), pp. 465-86; idem, Didymus the Blind, pp. 223-53. 

Modern Methods of Textual Criticism 239 

Method, which considers only the readings shared widely among 
group members (whether or not they are supported by members of 
other textual groups), the Comprehensive Profile Method considers 
a full range of readings, those found widely or universally within 
a group (irrespective of their occurrence elsewhere) and those 
found exclusively or principally among members of one group but 
not another. 

The method works best after a quantitative analysis has been 
performed and is meant to correct one of the problems of this older 
method: any member of a textual group should attest not only a high 
percentage of the readings of the group wherever there is genetically 
significant variation but also readings that are distinctive of the 
group, either because most (or all) of the group members otherwise 
attest it or because it is a reading attested by some (or all) group 
members but by members of no other group. 

The Comprehensive Profile Method proposes the use of three 
basic profiles to refine the findings of the quantitative analysis. The 
first profile (an "intergroup" profile) establishes the extent to which 
a newly collated witness attests readings found only among mem- 
bers of one group or practically only among members of one group; 
the second profile (an "intra-group" profile) establishes the extent to 
which the witness attests readings found in all other group members 
or among the majority of group members (irrespective of whether 
members of other groups also attest the reading); and the third pro- 
file combines the findings of the other two, establishing the extent to 
which a newly collated witness attests readings found universally or 
widely among members of one group but not among witnesses of 
the other groups. 

This kind of comprehensive evaluation of group readings can 
then refine the findings of ColwelPs quantitative analysis by provid- 
ing a definitive statement concerning the textual affinities of a newly 
collated witness. It has been used with particular success in analyses 
of the writings of important patristic sources. 71 

71. See Ehrman, Didymus the Blind; Darrell D. Hannah, The Text of 1 
Corinthians in the Writings ofOrigen (Atlanta, 1997); Roderic L. Mullen, 
The New Testament Text of Cyril of Jerusalem (Atlanta, 1997). The forthcom- 
ing volumes in the series are by John Brogan on Athanasius, Jean-Francois 
Racine on Basil the Great, and Carroll Osburn on Epiphanius. 

240 The Text of the New Testament 

VII. The Use of Computers in New 
Testament Textual Criticism 

The computer is beginning to revolutionize the field of textual 
studies. On the one hand, whereas texts of the New Testament 
circulated in manuscript form for some 1,500 years and in printed 
editions for some 500 years, they are now becoming increasingly 
available in electronic format. 72 Yet more important has been the use 
of the computer in furthering the traditional tasks of the discipline, 
establishing the original text of the New Testament and writing the 
history of its transmission. In particular, modes of data storage, re- 
trieval, and analysis that were barely conceivable 40 or 50 years ago 
are now becoming commonplace. 

It must be admitted that textual scholars were both slow and re- 
luctant to recognize the potential of computers for work in the 
field. 73 Some scholars were no doubt intimidated by the new tech- 
nologies; others were scornful of the idea that a computer would be 
able to make textual judgments concerning the "original" reading 
at a place of variation, when clearly such judgments must be made 
on grounds that require thought and (human) reasoning. But much 
of what textual critics do prior to decisions about textual variation 
involves data collection and comparison; this made it virtually in- 
evitable that critics should recognize the enormous potential for the 
computer in text-critical work, not as a substitute for human intuition 
and reasoning but as a tool to facilitate the endeavor. 

Initial attempts to utilize the computer for textual study date back 
to the 1950s, in the dissertation work of John Ellison at Harvard. 74 It 

72. See, e.g., the prototype of the edition critica maiorol the Institute 
for New Testament Textual Research at Miinster (http://nestlealand. 

73- See the opening words of Bonifatius Fischer, "Computer und der 
Text des Neuen Testamentes," in Studia I-vangelica, vi (Berlin, 1973), 
p. 109: "Wenn ein Neutestamentler heute von der Benutzung eines 
Computers redet, lauft er Gefahr, nicht mehr ernst genommen zu werden." 

74. Ellison's work was not completed until 1957, but it was begun 
already in 1950: "The Use of Electronic Computers in the Study of the 
Greek NT Text" (Diss., Boston, 1957). For a survey of the field from Ellison 
up to 1995, see Robert A. Kraft, "The Use of Computers in New Testament 
Textual Criticism," in The Text of the New Testament in Contemporary 
Research, ed. by Ehrman and Holmes, pp. 268-82. 

Modern Methods of Textual Criticism 241 

was not, however, until the 1970s that systematic steps began to be 
taken toward harnessing the power of the computer for textual work, 
for example, at the Institut fur neutestamentliche Textforschung in 
Miinster, under the leadership of Wilhelm Ott, and for the Vetus 
Latina project at Beuron, under Bonifatius Fischer. 75 These were still 
the days of computer punch cards and magnetic tape, however. 
Advances made then were simply preparatory for the explosion of 
technology that was to happen in the 1980s and 1990s. 

Even when the personal computer became the familiar working 
tool for virtually all scholars in the humanities, the use of computer 
technology was slow in coming into its own during the 1990s, as 
evident in a "state of the question" essay produced in 1995 by Robert 
Kraft, one of the leading figures in the field, who bemoaned the gen- 
eral lack of progress. 76 But the understandably pessimistic tone of 
Kraft's essay became dated nearly as soon as it was produced. For 
several computer initiatives (especially in relevant software) soon 
became more widely recognized and utilized, until today, when the 
major projects involving textual research (e.g., the International 
Greek New Testament Project [IGNTP] and the Institut fur neutesta- 
mentliche Textforschung [INTF]) now utilize computers extensively 
and with impressive results. 77 

Among all the developments in this field, the two most important 
involve aspects of the work that are foundational but painstaking and 

75. See, for example, Wilhelm Ott, "Transcription and Correction of 
Texts on Paper Tape: Experiences in Preparing the Latin Bible Text for the 
Computer," LASLA Revue, ii (1970), pp. 51-6; idem, "Computer Applications 
in Textual Criticism," in The Computer and Literary Studies, ed. by J. 
Aitken, et al. (Edinburgh, 1973), pp. 199-223; Bonifatius Fischer, "The Use 
of Computers in NT Studies, with Special Reference to Textual Criticism," 
Journal of 'theological Studies, n.s., xxi (1970), pp. 297-308; idem, "Com- 
puter und der Text," in Studia Evangelica. See also Vinton A. Dearing, 
"Determining Variations (in the Greek NT Text) by Computer," Society of 
Biblical Literature Seminar Papers, ii (1974), pp. 14-35; idem, Principles 
and Practice of Textual Analysis (Berkeley, 1974). 

76. Kraft, op. cit. 

77. See David C. Parker, "The Text of the Xew Testament and Com- 
puters: The International Greek New Testament Project," Literary and 
Linguistic Computing, xv (2000), pp. 27-41; Klaus Wachtel, "Editing the 
Greek New Testament on the Threshold of the Twenty-first Century," 
Literary and Linguistic Computing, xv (2000), pp. 43-50. 

242 The Text of the New Testament 

labor-intensive: collating manuscripts and presenting and analyzing 
the data derived from collations. 

The first is the COLLATE program devised by Peter Robinson for 
his study of the manuscript tradition of Chaucer. This program enables 
collators to enter their data (variations among the manuscripts) in such 
a way that the text of each manuscript is actually stored on a database, 
making it possible for the computer to calculate every place of varia- 
tion among every manuscript (it is capable of handling 100 manu- 
scripts). By customizing it for New Testament use, the program has 
significantly aided work undertaken by both the IGNTP and the INTF. 

The second is a program that was designed for the IGNTP by 
Jerry Lewis and Bruce Morrill, the latter a member of the project's 
North American committee. This program is specifically intended to 
facilitate the collation of witnesses of the Gospel of John by entering 
all the variations of a manuscript directly onto an existing database 
that contains both the base text and numerous places of previously 
noted variation. The program then allows the computer to calculate 
all the differences of all the manuscripts thus entered and to output 
the results in a range of formats deemed desirable by the user. 

Rather than explicate the precise workings of each of these pro- 
grams, which themselves continue to evolve and, in their current 
form, will no doubt soon become outdated, it is perhaps better to 
consider ways that they and other programs can assist in the tasks of 
textual criticism. 78 Four closely related and interlocking areas of ap- 
plication can be discussed. 

1. The Collection, Recording, 
and Storage of Data 

Before the advent of computers, textual data were gathered and 
recorded by hand. If a manuscript was to be collated, it was carefully 
compared, one letter at a time, with another manuscript or a base 
text (e.g., the Textus Receptus), and every (insignificant or signifi- 
cant) difference was noted on a sheet of paper. The collation would 
be published then as a list of variant readings from a base text. If it 
were to be used for an apparatus criticus, the collation would then 

78. For access to resources on the internet, see the frequently updated 
links on the NT Gateway webpage created and maintained by Mark 
Goodacre at the University of Birmingham ( 

Modern Methods of Textual Criticism 243 

need to be converted — that is, the variants from the base text (say, 
the Textus Receptus) would need to be reconfigured in relationship 
to the text of the critical edition being produced. At every stage — 
reading the manuscript, transcribing the collation by hand, convert- 
ing its data in relationship to a different base text, constructing the 
apparatus — there were possibilities of error. 

With new computer programs such as MANUSCRIPT and COL- 
LATE, the method of data recovery and storage is different. Already 
on the software program is a base text and variations to it that have 
been noted in previous collations. The collator then moves letter by 
letter through the manuscript being collated; wherever a difference is 
noted, it can be entered into the software's database. If it is a previously 
known variation, then the collator adds this manuscript's testimony 
to the list of manuscripts that have already been noted as attesting it. 
If it is a variation that has not previously been noted, the collator 
adds it to the database, indicating the number of the manuscript that 
now attests it. 

There are several key results obtained by this method of collation: 

1. It is much easier to check the accuracy of a collation. If two col- 
lators independently collate a manuscript electronically, the com- 
puter may be asked to compare the collations automatically; and 
where there are differences (e.g., where one collator notes a vari- 
ation but another does not), these can be called up and checked 
against the manuscript to verify the collation. 

2. Rather than presenting a collation simply as a list of variant readings 
from one particular base text, this method can allow each manu- 
script to be reconstructed completely in all its text. That is to say, 
the program takes the variant readings from the base text (and the 
lacunae that are noted, etc.) and automatically reconstructs the 
text of the witness itself so that what is produced is not only a set 
of variations, but an actual transcription of the manuscript. 

3- It is then possible to compare all of the collated witnesses (whose 
transcriptions are now stored in the database) against one an- 
other, simply by having the computer shift its base text to one of 
the witnesses and asking it to provide points of difference with 
any other witness or group of witnesses. 

This method of collation not only is faster, more accurate, and more 
useful, it also guarantees that the results of the collation are kept on 
permanent record. Once a witness is faithfully entered into the data- 
base, it never need be collated again. 

244 The Text of the New Testament 

2. The Presentation of the Data 

Before the development of electronic texts, editors were forced 
to make critical and non-reversible decisions about the format, pre- 
sentation, and extent of the text and apparatus. For example, if one 
wanted a critical text, then that text needed to be reconstructed (e.g., 
following the eclectic method) and an apparatus made in which 
some witnesses were represented but not others and some places 
of variation were noted but not others. These variations, of course, 
were always against the printed text, so an apparatus using the 
Textus Receptus as the base text (e.g., the IGNTP Luke volumes) 
would cite witnesses for different variations from one using the 
Nestle-Aland 27th edition. 

All of that has changed with the advent of electronic texts and 
databases that contain the results of collation. The computer itself 
can generate the apparatus from the variants included in its database, 
and it can do so against any text the user wants. Thus, for example, 
if a user wants to see the variant readings in John 18, the computer 
can be instructed (1) to present as the base text the text of any of the 
editions (e.g., Nestle-Aland, Westcott-Hort, Textus Receptus) or of 
any of the manuscripts that have been collated and (2) to list the 
variants from the selected text found in some or all other collated 
witnesses. For instance, if for some reason one wanted to use Codex 
Vaticanus or Codex Washingtonianus or the Textus Receptus as the 
base text, it would be a matter of hitting a few buttons and having it 
appear on the screen. When the base text is changed, so too, auto- 
matically, is the apparatus. 

Moreover, one can choose which manuscripts to include in the 
apparatus (e.g., just the Alexandrian witnesses, just the manuscripts 
of fam. 13, or every manuscript in the database). In addition, it is 
possible to include only certain kinds of variation in the apparatus 
(e.g., just omissions from the chosen base text, transpositions, or 
additions). What is more, the user can decide how variations are to 
be presented: they can come as a list of differences (as in most 
paper apparatus), or the actual text of each witness — say, for the 
entire verse — can be presented, with the variation in question 

In all of these options, the point is that the computer can quickly 
and accurately display the data that are wanted; one is not restricted 
to the data chosen by an editor in the evaluation of the textual 

Modern Methods of Textual Criticism 245 

3. Statistical Analyses 

Computers are particularly adept, of course, at doing complicated 
statistical analyses at blinding speed. Even as late as the early 1990s, 
if a scholar wanted to tabulate the percentage of agreements and 
disagreements of textual witnesses with one another (e.g., in order 
to do a quantitative analysis), 79 it had to be done by hand, each 
agreement and disagreement being carefully tabulated and counted 
and percentages determined by means of a hand calculator. Need- 
less to say, the process was painstaking and labor-intensive. 

With the refinement of computer technology, all of that is now 
completely unnecessary. The computer itself can churn out a com- 
plete list — or a partial list — of similarities and differences between 
two manuscripts or editions or, in fact, among any number of manu- 
scripts or editions. Moreover, for tabulations of results, the computer 
can be instructed concerning what kinds of variation to include (e.g., 
if it is thought desirable, as it usually is, not to count moveable v's and 
the alteration between ovrcog and ovtco) and can almost instanta- 
neously calculate the numbers of differences and the percentages of 
agreement among any witnesses whose transcriptions are available 
on the database. More sophisticated modes of analysis — for example, 
the Claremont Profile Method — can also be done electronically. 80 

These analyses in themselves will not resolve major textual 
cruxes — for example, concerning which reading is earlier or more 
likely "original" or how the history of transmission led from one type 
of text to another. For these cruxes to be solved, significant amounts 
of data need to be processed, and this can be done with incompara- 
bly greater speed and accuracy by computer than by hand. 

4. Hypertext Possibilities 

Before the appearance of the worldwide web and other elec- 
tronic modes of communication, scholars were severely restricted in 
their access to manuscripts, facsimile editions, published collations, 

79. See p. 234. 

80. For more sophisticated uses of computers to apply complicated 
statistical models for the classification of manuscripts, see J. C. Thorpe, 
"Multivariate Statistical Analysis for Manuscript Classification," T-C: A 
Journal of Biblical Textual Criticism ( vii (2002); Spencer, 
Wachtel, and Howe, op. cit., pars. 25-38. 

246 The Text of toe New Testament 

texts of the versions, collections of the writings of the Church fathers, 
and so on. Only scholars near a top-rate research library could have 
any hope of seeing firsthand a manuscript or any of the requisite 
repositories of data needed for the text-critical tasks. 

That too has changed with the development of computer tech- 
nology and promises to change yet more as this technology be- 
comes increasingly refined and popularized. It is now possible, of 
course, to digitize manuscripts and to put the writings of all the 
Church fathers in the original language on disk or to send the infor- 
mation over the internet. 

This will fling wide open the possibilities of electronic critical 
apparatus. One can imagine, very soon in fact, a web-based appara- 
tus that allows the user to click on a manuscript cited as containing 
a variant reading and have a transcription of the verse from the man- 
uscript appear or to click another link and have the image of the 
manuscript itself automatically come on the screen. The same would 
apply to citations of the versions and of the quotations of the Church 
fathers. One of the perennial problems with the latter is that editors 
who construct an apparatus must make judgments concerning 
whether a quotation of the New Testament is actually a citation or 
merely an allusion and whether the evidence can be accepted as cer- 
tain or not. With hypertext links it would be possible to click on the 
father's name in the apparatus and to be taken to the passage of his 
writings in question; this could appear in either the original language 
or in English translation. Also, it would be possible to look through 
the entire context so that the user can make a decision concerning 
the relevance of the citation for the text in question. 

David Parker has aptly stated that, with the construction of 
new electronic apparatus, "the study of the primary material will be 
democratized." 81 

VIII. Significant Ongoing Projects 

A number of text-critical projects are currently under way; among 
the most significant are those associated with the previously men- 
tioned INTF and the 1GNTP, headed by British and North American 

81. D. C. Parker, "Through a Screen Darkly: Digital Texts and the 
New Testament," Journal for the Study of the New Testament, xxv (2003), 
p. 409. 

Modern Methods of Textual Criticism 247 

1. The Institut fiir neutestamentliche 

En route to establishing its editio critica maior (see pp. 237-8), the 
INTF has made available a massive array of data in several subsidiary 
projects. The first is the multivolume series Text und Textwert der 
griechischen Handschriften des Neuen Testaments?' 2 The volumes of 
this series set out the results of the collations undertaken by the INTF 
in order to differentiate Greek manuscripts of the New Testament 
containing principally the Byzantine form of text from those that pre- 
serve other, earlier forms. The ultimate goal of the project is to ascer- 
tain which manuscripts need be cited individually in the apparatus 
criticus of the editio critica maior and which may be safely subsumed 
under the designation Byz. The volumes of Text und Textwert provide 
all of the data necessary for making these determinations. 

At the date of this writing (2004), the series has been completed for 
the general Epistles, the Pauline Epistles, the Book of Acts, and the Syn- 
optic Gospels. For each of these corpora, the volumes of the series pro- 
vide the following data: (1) descriptions of all of the available Greek 
manuscripts; (2) collations of all the manuscripts in the Teststellen ("test 
passages") chosen at the institute for determining which manuscripts 
support the Byzantine text, which support the "old form" of the text 
(taken to be the text represented in the editions of the United Bible 
Societies and Nestle-Aland Greek New Testament), and which support 
some other form of the text; and (3) a table to indicate, for each man- 
uscript, how many times it supports each form of the text. Once these 
data have been laid out, the volumes provide additional tables for those 
manuscripts that have a relatively high percentage of agreements with 
the "old form" of the text, indicating the number and percentage of 
agreements each one has with the 33 manuscripts that it agrees with 
most frequently in all of its readings (from among the Teststellen). 

These volumes cannot answer all the questions scholars may 
have concerning the relationship of the Greek manuscripts with one 
another since they are based only on "test passages," rather than full 
collations of each manuscript. For the Pauline Epistles, for example, 
the data are drawn entirely from 251 sets of variant readings found 
among the 742 manuscripts examined. This means that whereas the 
data presented are extensive — for the Pauline Epistles alone, they 

82. Published as volumes of the Arbeiten zur neutestamentlichen 
Textforschung (Berlin, 1987-1999). 

248 The Tkxt of the New Testament 

take up four large volumes (the first of which is 625 pages!) — they 
are necessarily incomplete as hundreds of other places of textual 
variation have not been considered. Even so, the series is valuable in 
removing all mystery concerning how the Institute has gone about 
deciding which manuscripts to cite explicitly in its apparatus criticus 
and which to consider predominantly Byzantine. 

A second, even more valuable series of publications by the Insti- 
tute is called Das Neue Testament auf Papyrus. 85 As its name sug- 
gests, the series is designed to lay out the text of each book of the 
New Testament as attested in the surviving papyri. Impressive in de- 
sign and execution, the volumes of the series present the complete 
texts of all known papyrus manuscripts of each New Testament book 
(to date, the Catholic and the Pauline Epistles have been covered). 
The editors provide full codicological and textual descriptions of 
each papyrus, along with bibliographical information on editions and 
secondary literature. The text of each papyrus, where extant, is then 
set forth in sublinear comparison with the running text of the Nestle- 
Aland Novum Testamentum Graece (which is given in extenso, i.e., 
even for portions of text not attested by the papyri). All textual dif- 
ferences, including spelling and abbreviations, are indicated for each 
papyrus; a complete apparatus notes correctors, lacunae, and earlier 
reconstructions. A second apparatus delineates every textual variation 
uncovered in fresh and verified collations of all the Greek majuscule 
manuscripts for every verse of the epistles, including those for which 
no papyrus is yet extant. In terms of the number of variant readings, 
this apparatus is far more extensive than the Nestle-Aland apparatus 
criticus; between text and apparatus, one can deduce the complete 
manuscript complexion of the first ten centuries. 

2. The International Greek New Testament Project 

Among the by-products of the efforts of the IGNTP to produce an edi- 
tion, with lengthy apparatus, of the fourth Gospel, 84 two are especially 
worth noting. The first is a two-volume work edited by David Parker 
and William Elliott that provides exhaustive access to (1) the texts 
of all of the 23 surviving papyri of John 85 and (2) the 28 complete and 

83. Also published as volumes of Arbeiten zur nentestamentliche 
Textforschung (Berlin, 1986-1994). 

84. See p. 192, n. 43. 

85. The New Testament in Creek IV: 'The Gospel According to St. John. 
vol. i, The Papyri, ed. by W. J. Elliott and D. C. Parker {New Testament 
Tools and Studies, xx; Leiden, 1995). 

Modern Methods of Textual Criticism 249 

36 fragmentary majuscule manuscripts of John. 86 Each volume consists 
of four major parts: (1) description of the witnesses, including bibli- 
ography of secondary studies and opinions concerning dates; (2) com- 
plete transcriptions of each witness; 87 (3) an apparatus criticus, which 
indicates all the variations (including spelling differences) of each 
witness against a base text (the 1873 Oxford edition of the Textus 
Receptus); and (4) a set of photographic plates of each witness. 88 
By providing transcriptions, apparatus critici, and photographs, the 
editors of these volumes have given their readers complete and reli- 
able access to these important witnesses to the fourth Gospel. 

The second by-product of the IGNTP's efforts to establish a com- 
prehensive apparatus for the fourth Gospel is the Byzantine Text Proj- 
ect, headed by David Parker and directed by Rod Mullen. The project 
began in response to a request from a consultation of Eastern Ortho- 
dox scholars meeting in El Escorial, Spain, in 1999 to the United Bible 
Societies to "consider producing a critical edition of the Byzantine 
text." Unlike other recent "majority text editions," such as those of 
Hodges/Farstad and Pierpont/Robinson, 89 the Byzantine Text Project 
does not aim to reconstruct a close approximation of the "original" text; 
its purpose is to show the rich variety of the Byzantine textual tradition 
by providing an apparatus that includes manuscripts from all of the im- 
portant subgroups and families within that tradition, making it possible 
for users to determine which form of the text was widely available at 
different times and places of its transmission. As with the IGNTP vol- 
umes on Luke, therefore, the emphasis will be placed not on the text 
that is printed (in this case, it will use the minuscule manuscript 35 as 
its base text) but on the apparatus. The project anticipates having an 
edition of the fourth Gospel completed by the end of 2004. 

86. The New Testament in Greek IV, volume ii, The Majuscules 
(expected in 2004). 

87. The papyrus volume does not include transcriptions of the longer 
papyri, p 66 and p 75 , as these are already available to scholars in the publica- 
tions of the Bodmer papyri collections; see pp. 56-58. So too the majus- 
cule volume provides transcriptions only for the 36 fragmentary witnesses, 
although the readings of the 28 complete majuscules are included in the 
apparatus criticus. 

88. For all the papyri except, again, p 66 and p 75 (see n. 87), these 
photos are complete; for others, representative pages are included; in the 
majuscule volume, complete photographs of the fragmentary manuscripts 
and sample pages from the complete manuscripts are provided. 

89. See pp. 219-20. 


The Causes of Error in the 

Transmission of the Text 

of the New Testament 

As the physician must make a correct diagnosis of a disease before 
attempting to effect its cure, so the textual critic must be aware of the 
several kinds of injury and danger to which a text transmitted by 
handwriting is liable to be exposed before trying to rectify the errors. 
In fact, it is important to see not only what might happen but also 
what has happened in the copying of manuscripts. No systematic 
attempt is made in this chapter to evaluate the relative worth of the 
variant readings; the purpose is to describe and classify the phe- 
nomena rather than at this point to prescribe the remedy. 1 

1 . For a discussion of the subject from an older and different point of 
view, see John W. Burgon, The Causes of the Corruption of the Traditional 
Text of the Holy Gospels, ed. by Edward Miller (London, 1896). A suggestive 
analysis of the origin of errors in terms of the several stages involved in 
the process of copying is given by Eugene Vinaver, "Principles of Textual 
Emendation," Studies in French Language and Mediaeval Literature 
Presented to Prof. Mildred K. Pope (Manchester, 1939), pp. 351-69; see also 
J. Andrieu, "Pour Pexplication psychologique des fautes de copiste," Revue 
des etudes latines, xxviii (1950), pp. 279-92. Other studies of scribal prac- 
tices and habits include the following: Klaus Junack, "Abschreibpraktiken 
und Schreibergewohnheiten in ihrer Auswirkung auf die Textuberlieferung," 


Causes of Error in Transmission of the Text 251 

I. Unintentional Changes 

1 . Errors Arising from Faulty Eyesight 2 

The scribe who was afflicted with astigmatism found it difficult 
to distinguish between Greek letters that resemble one another, par- 
ticularly when the previous copyist had not written with care. Thus, 
in majuscule script, the sigma (which was customarily made as a 
lunar sigma), the epsilon, the theta, and the omicron (C, €, 0, O) 
were sometimes confused. For example, in Acts 20.35, three minus- 
cule manuscripts (614, 1611, and 2138) read xomcbvra edei instead 
of xomtivxac, del, an error that goes back to a majuscule ancestor 
written in scriptio continua. In 1 Tim. 3-16, the earlier manuscripts 
read OC ("he who") while many of the later manuscripts read 0C 
(the usual contraction for Oedg, "God"). The letters gamma, pi, and 
tau (r, fl, T) were liable to be confused, particularly if the crossbar 
on the first and last letters had been carelessly drawn or if the right 
leg of the pi was too short. Thus, in 2 Pet. 2.13, some manuscripts 
read AEA.IIAI2 ("love feasts") and others read AFLATAE; ("decep- 
tions"). If two lambdas were written too close together, they could 
be taken as the letter mu, as has happened at Rom. 6.5, where most 
manuscripts have AAAA ("but") but others have AMA ("together"). 
If a lambda is followed too closely by an iota, the combination may 
look like the letter nu (AI and N). Thus, in manuscripts of 2 Pet. 
2.18, OAirQS ("scarcely") alternates with ONTQ2 ("really"), where 
the tau and the gamma are also confused. Sometimes A and A were 
mistaken for each other, as in Acts 15.40, where EITIAEHAMENOS 
("having chosen") appears in Codex Bezae as EIIIAEHAMENOS 
("having received"). 

in New Testament Textual Criticism: Its Significance for Exegesis. Essays in 
Honour of Bruce M. Metzger, ed. by Eldon J. Epp and Gordon D. Fee 
(Oxford, 1981), pp. 277-95; James Royse, "Scribal Tendencies in the Trans- 
mission of the Text of the New Testament," in The New Testament in Contem- 
porary Research: Essays on the Status Quaestionis, ed. by Bart D. Ehrman and 
Michael W. Holmes (Grand Rapids, MI, 1995), pp. 239-52; P. M. Head and 
M. Warren, "Re-inking the Pen: Evidence from P. Oxy. 657 (P 13 ) Concerning 
Unintentional Scribal Errors," New Testament Studies, xliii (1997), pp. 466-73. 
2. According to Edward Rosen, eyeglasses were invented in Italy dur- 
ing the early part of the fourteenth century, see his article, "The Invention 
of Eyeglasses," Journal of the History of Medicine and Allied Sciences, 
vol. 11 (1956), pp. 13^46 and 183-218. 

252 The Text of the New Testament 

In the generally accepted text of 1 Cor. 12.13, Paul declares "By 
one Spirit we were all baptized into one body . . . and all were 
made to drink of one Spirit." Several witnesses, however, conclude 
the statement thus: "all were made to drink of one drink," a variant 
that arose when scribes misread the letters FIMA (the usual contrac- 
tion of the word nvevfia) as FIOMA ("drink"). Since the word xai 
was sometimes abbreviated K,, a kappa with a heavy dot of ink at the 
end of the lower diagonal line might be taken as the syllable xai. 
This has in fact happened at Rom. 12.11, where the curious variant 
"serving the time" (ra> xmgd) dovXevovxeq) arose from the true text, 
"serving the Lord" {rqj xvgiq) dovXevovTeg), because a scribe took 
the contraction of the word xvquo (KQ) as KPQ. 

As was mentioned in the first chapter (see p. 17), in antiquity 
nonliterary, everyday documents were customarily written in a 
cursive hand in which most of the letters were formed without lifting 
the pen and abbreviations were freely used. Whether any of the 
books of the Greek Bible ever circulated in cursive or semicursive 
script is a question to which different answers have been given. 
Wikenhauser, following Roller, argued that it is unlikely that the orig- 
inal texts of the New Testament books were written in cursive 
script. 3 Several years later, however, in the newly expanded format 
of this standard work, Josef Schmid declared that the New Testament 
authors would have undoubtedly ("zweifellos") used the cursive 
style of writing. In support of this view, it was pointed out by Nau 4 
that in the books of Chronicles Codex Vaticanus contains certain per- 
mutations of the letters mu, nu, and beta that cannot be explained in 
terms of confusion of majuscule script, where these three letters are 
very different from one another, but are readily explainable in cur- 
sive script, where they resemble each other closely. For example, the 
permutation of n and v in 2 Chron. 16.7, Ava/iei for Avavei; 17.8, 
Mavdaviaq and Icogav for NavOaviag and Icogafi; 31 12-13, 
XtDjueviag and MaeO for Xtoveviag and NaeO; the permutation of /5 
and v in 2 Chron. 17.8, Tcojiadcofieia for Tcvfiadcoveia; and the per- 
mutation of /? and fi in 2 Chron. 21.10, Aojiva for Aofiva; 36.2, 
Afieixak for AfieiraX. 

3. Alfred Wikenhauser, Einleitung in das Neue Testament (Freiburg, 
1953), pp. 49-50; English trans. New Testament Introduction (New York, 
1958), p. 67; and Wikenhauser and Josef Schmid, Einleitung in das Neue 
Testament, sechste, vollig neu bearbeitete Aufl. (Freiburg im Breisgau, 
1973), p. 69. 

4. See F. Nau in Revue de I'orient Chretien, xvi (1911), pp. 428-9. 

Causes of Error in Transmission of the Text 253 

Another example of a biblical manuscript that undoubtedly goes 
back to a cursive ancestor is the Berlin fragment of Genesis, a 
papyrus copy in semicursive script dating from the third century A.D. 
From a study of a wide variety of scribal errors in the text, the editors 
concluded that one or more ancestors were written in a typical 
cursive hand. 5 

When two lines in the exemplar being copied happened to end 
with the same word or words, or even sometimes with the same 
syllable, the scribe's eye might wander from the first to the second, 
accidentally omitting the whole passage lying between them. Thus is 
to be explained the curious reading at John 17.15 in Codex Vaticanus, 
which lacks the words that are enclosed in square brackets: "I do not 
pray that you take them from the [world, but that you keep them 
from the] evil one." In the exemplar from which the scribe of this 
manuscript was copying, the Greek text probably stood in the 
following arrangement: 

avrovg EX TOV 


avrovg ex xov 


After copying the first line, the scribe's eye returned, not to the be- 
ginning of line 2, but to the beginning of line 4. Such an error is 
called parablepsis (a looking by the side) 6 and is facilitated by 
homoeoteleuton (a similar ending of lines). 

5. See the list of scribal errors collected by Henry A. Sanders and Carl 
Schmidt, The Minor Prophets in the Freer Collection and the Berlin Fragment 
of Genesis (New York, 1927), pp. 244-6. For examples of variant readings in 
Codex Bezae that may have arisen in a cursive ancestor, see Paul Glaue's lists 
in Zeitschrift fur die neutestamentliche Wissenschaft, xlv (1954), pp. 92—4. 

It is possible that p 7S had a cursive ancestor, for in Luke 16.19 the 
name of the rich man is (ovo^an) Nsvng, which is thought to stand by 
haplography for (ovo/tari) Ntvevnc.. Now, though the similarity of n and vi 
is not close when written in majuscules, in some forms of first-century cur- 
sive script, the syllable zi could have been very easily confused with the 
letter v; for examples, see E. M. Thompson, Handbook of Greek and Latin 
Palaeography (New York, 1893), chart facing p. 148; L. Gonzaga da 
Fonseca, Epitome introductions in palaeographiam graecam (biblicam), 
ed. altera (Rome, 1944), pp. 85, 94 f. 

6. On the possibility of the scribe's eye wandering to a different col- 
umn of the exemplar, see the examples cited by J. Rendel Harris in the 
American Journal of Philology, vi (1885), pp. 25-40. 

254 The Text oe the New Testament 

Many other examples of omission, called haplography, occur in 
a wide variety of manuscripts. For example, the whole verse at Luke 
10.32 is lacking in K because the sentence ends with the same verb 
(dvri7taQf}X9ev) as the previous sentence (verse 3D. Codex Alexan- 
drinus omits the entire verse at 1 Cor. 9.2, which ends with the same 
four words (v/tEig iaze ev xvqioj) as the previous verse. Since the last 
five words in Luke 14.26 and 27 are exactly the same (ov dvvazm 
eivaifiov fiaOrjrrjg), it is easy to account for the accidental omission 
of verse 27 in more than a dozen different manuscripts. The words 
of 1 John 2.23, "He who confesses the Son has the Father also," fell 
out of the later manuscripts (on which the King James Version de- 
pends) because of the presence of rov Jtarega fyst in an adjacent 
clause. Other interesting cases of error arising from homoeoteleuton 
are found in various manuscripts at Luke 5.26, 11.32, and 12.9 and at 
Rev. 9.2-3. 

Sometimes the eye of the scribe picked up the same word or 
group of words a second time and, as a result, copied twice what 
should have appeared only once (this kind of error is called dittog- 
raphy). In Acts 19.34, the cry of the mob, "Great is Artemis of the 
Ephesians," is given twice in Codex Vaticanus. Again, instead of the 
generally accepted text of Acts 27.37, "We were in all two hundred 
and seventy-six (cog) 7 persons in the ship (ev xq> jiXoico)," Codex 
Vaticanus and the Sahidic version read " about seventy-six bbgog)." 
The difference in Greek is slight (FIAOIQCOS and IIAOIQQCOS). 

2. Errors Arising from Faulty Hearing 

When scribes made copies from dictation or even when a soli- 
tary scribe pronounced aloud the words being transcribed, confusion 
would sometimes arise over words having the same pronunciation as 

7. After the second century B.C., the letters of the Greek alphabet 
served as numerals. In addition to the 24 letters of the alphabet, three 
obsolete signs were employed: the digamma (f) or stigma (g) for 6, the 
koppa (0 or O for 90, and the sampi (^\) for 900. The first nine letters of 
the alphabet stand for units, the second nine for tens, the third nine for 
hundreds: d = 1, /?' = 2, y = 3, 6' = 4. e' = 5, for g' = 6, £' = 7, 
rj' =8,6'= 9, 1 = 10 (id = 11, t/?' = 12, etc.), x = 20 (xd = 21, etc.), 
A' = 30, n' = 40, v = 50, I' = 60, d = 70, Jt' = 80, o' or C = 90, 
q' = 100, a = 200, x = 300, v = 400, <p' = 500, % = 600, y>' = 700, 
a)' = 800, ^' = 900, ,a = 1,000, ,0 = 2,000, ,y = 3,000, etc. 

Causes of Error in Transmission of the Text 255 

others but differing in spelling (as the English words there and their 
or grate and great). During the early centuries of the Christian era, 
certain vowels and diphthongs of the Greek language lost their dis- 
tinctive sounds and came to be pronounced alike, as they are today 
in modern Greek. The confusion between a> and o was common, 
accounting for such variants as EXOfisv and E%OfiEV in Rom. 5.1 and 
wbe and ode in Luke 16.25. 

The diphthong ai and the vowel £ came to be pronounced alike 
(with a short e sound). As a result, the second-person plural ending 
-ode sounded the same as the ending of the middle and passive 
infinitive -odai, accounting for the variants EQXEoOai and EQ%eo6e in 
Luke 14.17, ^rjlovade and tyXovodcu in Gal. 4.18, and similarly 
in many other passages. Sometimes the change of vowels resulted in 
an entirely different word. Thus, in Matt. 11.16 EtEQOig ("others") in 
some manuscripts varies with Exaiooig ("comrades") in others. 

The pronunciation of ov and of v was sometimes indistinguish- 
able and accounts for the variation in Rev. 1.5. The translators of the 
King James Version followed a text of this verse that had Xovoavxi 
("Unto him that loved us, and washed us from our sins in his own 
blood"), whereas the text used by modern translators reads the verb 
Xvaavxi ("and freed us from"), which is found in the earlier Greek 

In Koine Greek, the vowels t], i, and v, the diphthongs ei, oi, 
and vi; and the improper diphthong y came to be pronounced 
alike, all of them sounding like ee in the English feet. It is not 
surprising that one of the commonest kinds of scribal confusion in- 
volves the substitution of these seven vowels or diphthongs for 
one another. This kind of error, which is commonly called itacism, 
accounts for several extremely odd mistakes present in otherwise 
good manuscripts. For example, in 1 Cor. 15.54, the statement 
"Death is swallowed up in victory (vlxog)" appears in p 46 and B 
as "Death is swallowed up in conflict (vslxog)." According to the 
vision of the seer on Patmos, around God's throne in heaven was "a 
rainbow (.igig) that looked like an emerald" (Rev. 4.3). In K, A, and 
other witnesses, one finds the similarly pronounced word "priests" 

In view of the ever-present possibility of committing itacism, it is 
not surprising that the evidence for the Greek personal pronouns 
varies widely in New Testament manuscripts (as f)/u£ig/vfi£ig, 
fjfitv/vfilv, and fjfidg/v/idg). Problems arise especially in the episto- 
lary literature. Did John write his first letter "that our (fi/iajv) joy may 

256 The Text of the New Testament 

be complete" or "that your (v/icov) joy may be complete"? Does Paul 
include himself with his readers at Gal. 4.28 by using fjjuelg, or did 
he write fifiEig? Whichever reading is judged to be original, it is easy 
to see how the other arose. In the five chapters of 1 Peter, the man- 
uscripts contain at least seven instances of such an interchange of 
personal pronouns (1.3, 12; 2.21 [twice]; 3.18, 21; 5.10). Occasionally 
a confusion of personal pronouns took place that produced virtual 
nonsense in the context; Paul's solemn statement in 2 Thess. 2.14, 
"He called you (vjudg) through our gospel, so that you may obtain 
the glory of our Lord Jesus Christ," reads in manuscripts A, B, D*, 
1881, "He called us (ffftag) through our gospel ..." So widespread is 
this kind of scribal error that the testimony of even the best manu- 
scripts respecting personal pronouns is liable to suspicion, and one's 
decision between such variant readings must turn upon considera- 
tions of fitness in the context. 

Besides eliminating differences in the pronunciation of certain 
vowel sounds, later Greek ceased to give the rough breathing a dis- 
tinctive force. Manuscripts that have rough and smooth breathing 
marks often use them most arbitrarily so that avzov varies with 
avxov, earrjxev (perfect tense of ioxtj/m) varies with eottjkev (imper- 
fect tense of onjxcj), etc with elg, and other similar pairs of words. 8 

In addition to confusion of vowels that sounded alike, certain 
consonants are occasionally interchanged, as in Matt. 2.6 ex oov 
("from you") becomes i £ ov ("from whom") in K c (see also Matt. 
21.19 and Mark 11.14). In the same category belong instances of con- 
fusion between forms of verbs spelled with a single or double con- 
sonant (e.g., the present and the second aorist stems epceXXev and 
EfieXev in John 12.6) and the confusion of different words altogether 
(e.g., EyEvvrjdrjaav and iyEvrjOrjoav in John 1.13)- Somewhat simi- 
lar are the readings in 1 Thess. 2.7, where the pronunciation of 
ey£Vtj8r}fi£V rjmoi ("we were gentle") is almost indistinguishable 
from that of iyevrjOrjfiEv vijjuol ("we were babes," see pp. 328-30). 

A curious interchange of consonants has taken place at 
Rev. 15.6, where the description of the seven angels as "robed in 
pure bright linen" (Aivov) becomes "robed in pure bright stone" 
(kLOov) in several early manuscripts (including A, C, and codices of 

8. For other examples with references to specific New Testament 
manuscripts, see Blass, Debrunner, and Funk, A Greek Grammar of the 
New Testament and Other Early Christian Literature (Chicago, 1961), § 14. 

Causes of Error in Transmission of the Text 257 

the Vulgate). At Heb. 4.11, the scribe of Codex Claromontanus wrote 
dXijdeiag ("truth") for cuieiBeiag ("disobedience"), with quite disas- 
trous results to the sense! 

3. Errors of the Mind 

The category of errors of the mind includes those variations that 
seem to have arisen while the copyist was attempting to hold a 
clause or a sequence of letters in a somewhat treacherous memory 
between glancing at the manuscript being copied and writing down 
what had been seen there. In this way, one must account for the 
origin of a multitude of changes involving substitution of synonyms, 
variation in word order, and transposition of letters. 

1. Substitution of synonyms may be illustrated by the following 
examples: ebtsv for E(j>rf, ex for cam and the reverse, evBvc; for evdeajc; 
and the reverse, on for dtori, tieqi for ineg and the reverse, and 
dfipidxwv for 6<p6aAjud)v. 

2. Variation in the sequence of words is a common phenome- 
non; thus, the three words Jidvzeg xal Epanxi^ovxo in Mark 1.5 also 
appear in the order xal tfiajixiCpvxo ndvxeq as well as xal jtdvxeg 
efianxiCpvxo . 

3. Transposition of letters within a word sometimes results in the 
formation of a different word, as eAafiov in Mark 14.65 becomes 
efiaXov in some manuscripts (and efialXov in other manuscripts). 
Such alterations of letters sometimes produce utter nonsense; at John 
5.39, where Jesus speaks of the Scriptures as "they that bear witness 
(ai fiaoxvaovaai) concerning me," the scribe of Codex Bezae wrote 
"they are sinning (afiaoxdvovoai) concerning me"! 

4. Assimilation of the wording of one passage to the slightly 
different wording in a parallel passage, which may have been bet- 
ter known to the scribe, accounts for many alterations in the Syn- 
optic Gospels. Thus, at Matt. 19.17, the reading of the earlier man- 
uscripts, "Why do you ask me about what is good? One there is 
who is good," is changed in later manuscripts to agree with the 
form of Jesus' words as reported in Mark 10.17 and Luke 18.18, 
"Why do you call me good? No one is good but God alone." (The 
King James translators followed the later form of text in Matthew.) 
More than once in the Epistles to the Colossians and to the 
Ephesians scribes have introduced into passages of one epistle 
words and phrases that properly belong to parallel passages in the 

258 The Text of the New Testament 

other. Thus, to the statement in Col. 1.14 "in whom we have 
redemption, the forgiveness of sins," a few later Greek manuscripts 
add the words "through his blood," a phrase derived from the par- 
allel in Eph. 1.7. (Here again the King James Version follows the 
secondary form of text.) 

4. Errors of Judgment 

Though perhaps several of the following examples might be 
classified under the category of deliberate changes introduced for 
doctrinal reasons, it is possible to regard them as unintentional errors 
committed by well-meaning but sometimes stupid or sleepy scribes. 

Words and notes standing in the margin of the older copy were 
occasionally incorporated into the text of the new manuscript. Since 
the margin was used for glosses (i.e., synonyms of hard words in the 
text) as well as corrections, it must have often been most perplexing 
to a scribe to decide what to do with a marginal note. It was easiest 
to solve any doubt by putting the note into the text being copied. 
Thus, it is probable that what was originally a marginal comment ex- 
plaining the moving of the water in the pool at Bethesda (John 5.7) 
was incorporated into the text of John 5.3b-4 (see the King James 
Version for the addition). Again, it is altogether likely that the clause 
in later manuscripts at Rom. 8.1 "who walk not according to the flesh 
but according to the spirit" was originally an explanatory note (per- 
haps derived from verse 4) denning "those who are in Christ Jesus." 
As was mentioned in Chapter 1, some manuscripts are provided with 
marginal helps designed to assist the reader of the fixed Scripture 
lessons appointed by the ecclesiastical calendar (the Lectionary). As 
a result, lectionary formulas, such as ebiev b xvgiog, occasionally 
crept into the text of nonlectionary manuscripts (e.g., at Matt. 25.31 
and Luke 7.3D. 9 

Other errors originated, not because of the exercise of faulty judg- 
ment, but from the lack of judgment altogether. Only heedlessness to a 
degree that passes comprehension can account for some of the absur- 
dities perpetrated by witless scribes. For example, after sig rovg dylovg 
of 2 Cor. 8.4 a good many minuscule manuscripts have added the gloss 
SegaoOai f/fidg. It appears that a scribe of one of these manuscripts 
wrote in the margin beside ddgaoOcu fjfJ-ac; the comment ev nokXoZg 

9. For other examples of influence from lectionaries, reference may 
be made to B. M. Metzger, The Saturday and Sunday Lessons from Luke in 
the Greek Gospel Lectionary (Chicago, 1944), pp. 14-17. 

Causes of Error in Transmission of the Text 259 

x&v dvTiyQa<j)wv ovrcog evgrjTai ("it is found thus in many of the 
copies"). The scribe of a subsequent manuscript (cited by Bengel) 
incorporated this comment on the gloss directly in the text as though it 
were part of the apostle Paul's instructions to the Corinthians! 10 

What is perhaps the most atrocious of all scribal blunders is con- 
tained in the fourteenth-century Codex 109- This manuscript of the 
four Gospels, now in the British Library, was transcribed from a copy 
that must have had Luke's genealogy of Jesus (3.23-38) in two 
columns of 28 lines each. Instead of transcribing the text by follow- 
ing the columns in succession, the scribe copied the genealogy by 
following the lines across the two columns. 11 As a result, not only is 
almost everyone made the son of the wrong father but, because the 
names apparently did not fill the last column of the exemplar, the 
name of God now stands within the list instead of at its close (it 
should end, of course, "Adam, the son of God"). In this manuscript, 
God is actually said to have been the son of Aram and the source of 
the whole race is not God but Phares! 

II. Intentional Changes 12 

Odd though it may seem, scribes who thought were more dan- 
gerous than those who wished merely to be faithful in copying what 
lay before them. Many of the alterations that may be classified as 

10. For a list of more than 600 such instances, see Reuben J. Swanson, 
"Unique and even Bizarre Readings in Manuscripts of 1 Corinthians," New 
Testament Greek Manuscripts . . . 1 Corinthians (Wheaton, IL, 2003), 

pp. 459-87. Swanson also supplies a list of "Passages Marked with Umlaut ["] 
in Vaticanus," pp. 293-304. See also J. Edward Miller, "Some Observations 
on the Text-Critical Function of the Umlauts in Vaticanus, with Special At- 
tention to 1 Corinthians 14.34-35," Journal for the Study of the New 
Testament, xxvi (2003), pp. 217-36. Miller concludes his discussion: 

The Vaticanus scribe consistently positions the umlaut beside an un- 
certain line of text rather than the line preceding it. This eliminates 
1 Cor. 14.34-35 as the variant indicated by the nearby umlaut, since 
the umlaut marks the line containing the final words of 1 Cor. 14.33 
and not the following line which begins with 1 Cor. 14.34. 

11. For a description of other manuscripts in which the Lucan genealogy 
of Jesus is confused to a greater or lesser extent, see Jacob Geerlings, Family 
TI in Luke {Studies and Documents, xxii, Salt Lake City, UT, 1962), pp. 127-37. 

12. For other discussions of this subject, see Eric L. Titus, "The Moti- 
vation of Changes Made in the New Testament Text by Justin Martyr and 

260 Thf. Text of the New Testament 

intentional were no doubt introduced in good faith by copyists who 
believed that they were correcting an error or infelicity of language 
that had previously crept into the sacred text and needed to be rec- 
tified. 13 A later scribe might even reintroduce an erroneous reading 
that had been previously corrected. For example, in the margin of 
Codex Vaticanus at Heb. 1.3 there is a curiously indignant note by a 
rather recent scribe 14 who restored the original reading of the codex, 
(j}avEQWv, for which a corrector had substituted the usual reading, 
<p£Q(ov: "Fool and knave, leave the old reading, don't change it!" 
{dfiaGearare xai nani, &<peg xbv nakavov, fit] fieTcuioiei). 

Clement of Alexandria: A Study in the Origin of New Testament Variation" 
(Diss., Chicago, 1942); C. S. C. Williams, Alterations to the Text of the Synop- 
tic Gospels and Acts (Oxford, 1951); Leon E. Wright, Alterations of the Words 
of Jesus as Quoted in the literature of the Second Century (Cambridge, MA, 
1952); E. W. Saunders, "Studies in Doctrinal Influence on the Byzantine Text 
of the Gospels," Journal of Biblical Literature, Ixxi (1952), pp. 85-92; K. W. 
Clark, "Textual Criticism and Doctrine," in Studia Paulina in honorem 
Johannis de Zwaan (Haarlem, 1953), pp. 52-65; Eric Fascher, Textgeschichte 
als hermeneutisches Problem (Halle/S., 1953); Manfred Karnetzki, 
"Textgeschichte als Oberlieferungsgeschichte," Zeitschrift fur die neutesta- 
mentliche Wissenschaft, xlvii (1956), pp. 170-180; Eldon Jay Epp, The Theo- 
logical Tendency of Codex Bezae Cantabrigiensis in Acts (Cambridge, 1966); 
Alexander Globe, "Some Doctrinal Variants in Matthew 1 and Luke 2 and 
the Authority of the Neutral Text," Catholic Biblical Quarterly, xlii (1980), 
pp. 52-72; Peter M. Head, "Christology and Textual Transmission: Reveren- 
tial Alterations in the Synoptic Gospels," Novum Testamentum, xxxv (1993), 
pp. 105-29; Bart D. Ehrman, The Orthodox Corruption of Scripture: The 
Effect of Early Christological Controversies on the Text of the New Testament 
(Oxford, 1993); idem, "The Text as Window: New Testament Manuscripts 
and the Social History of Early Christianity," in The Text of the New Testa- 
ment in Contemporary Research, ed. by Ehrman and Holmes, pp. 361-79- 

13. Jerome complained of the copyists who "write down not what 
they find but what they think is the meaning; and while they attempt to 
rectify the errors of others, they merely expose their own" (scribunt non 
quod inveniunt, sed quod intellegunt; et dum alienos errores emendare 
nituntur, ostendunt suos), Epist. 71.5, Ad LucinumQ. P. Migne, Patrologia 
Latina, 22,671; Corpus Scriptorum Eccleriasticorum Latinorum, 55, pp. 5 f.). 
For other patristic references to the incompetence of some copyists, see 
pp. 101 and 152. 

14. The scribe was perhaps of the thirteenth century; see the plate in 
Bruce M. Metzger, Manuscripts of the Greek Bible: An Introduction to Greek 
Palaeography (New York, 1981), pp. 74-5. 

Causes of Error in Transmission of the Text 261 

Andrew of Caesarea in Cappadocia, in his commentary on the 
Book of Revelation, 15 written about the year 600, expressly applied 
the curse recorded in Rev. 22.18-19 to those litterati who considered 
that Attic usage 16 and a strictly logical train of thought were more 
worthy of respect and more to be admired (afyomoTOXEQa aai 
oe/ivorsga) than the peculiarities of biblical language. What Andrew 
refers to is illustrated by an anecdote told by Sozomen, a fifth- 
century lawyer of Constantinople who wrote a history of the Church. 
He relates that at an assembly of Cypriot bishops about the year 350, 
one Triphyllios of Ledra, a man of culture and eloquence, was ad- 
dressing the assembly and, in quoting the text "Rise, take up your 
bed and walk," substituted the more refined Attic Greek word 
oxifinovQ for the colloquial Koine word XQd^j3arog ("pallet") used 
in John 5.8. Whereupon a certain Bishop Spyridon sprang up and in- 
dignantly called to him before the whole assembly, "Are you, then, 
better than he [Jesus] who uttered the word XQafifiaTog, that you are 
ashamed to use his word?" 17 Despite the vigilance of ecclesiastics of 
Bishop Spyridon's temperament, it is apparent from even a casual 
examination of a critical apparatus that scribes, offended by real or 
imagined errors of spelling, grammar, and historical fact, deliberately 
introduced changes into what they were transcribing. 

1 . Changes Involving Spelling and Grammar 

The Book of Revelation, with its frequent Semitisms and sole- 
cisms, afforded many temptations to style-conscious scribes. It is not 
difficult to imagine, for example, that the use of the nominative case 
after the preposition and (in the stereotyped expression, cued 6 wv 

15. Josef Schmid, Studien zur Geschichte des griechiscben Apokalypse- 
Textes; i, Der Apokalypse-Kommentar des Andreas von Kaisareia (Munich, 
1955), p. 262, 11. 3-12. 

16. Wilhelm Michaelis, "Der Attizismus und das Neue Testament," 
Zeitschrift fur die neutestamentliche Wissenschaft, xxii (1923), pp. 91-121; 
G. D. Kilpatrick, "Atticism and the Text of the Greek New Testament," 
repr. in The Principles and Practice of New Testament Textual Criticism: 
Collected Essays of G. D. Kilpatrick, ed. by J. K. Elliott (Leuven, 1990), 

pp. 15-32. 

17. Sozomen, Hist. Keel. 1.11. Eusebius tells us that Tatian ventured to 
paraphrase certain words of the apostle Paul "as though improving their style" 
((bgemdioQOov/ievovavrdivrfjg^QdoecDgovvra^iv), Hist. Eccl. 4.39.6. 

262 The Text of the New Testament 

ml 6 rjv xal 6 igxdfievog, Rev. 1.4) would grate on the sensibilities 
of Greek copyists and that, consequently, they would insert after 
and either tov, 9eov, or xvqlov in order to alleviate the syntax. As 
a matter of fact, all three of these attempts to patch up the grammar 
are represented today in one or more manuscripts. 

The use of xai joining the finite verb ijioirjoev in Rev. 1.6 to the 
participles in 1.5 strains the rules of Greek concord beyond the 
breaking point; scribes mended the syntax by changing the indica- 
tive to another participle (jioirjoavn). The genitive case of 
jisjiVQCOfiivrjg in Rev. 1.15, which agrees with nothing in its clause, 
was altered by some scribes to the dative and by others to the nom- 
inative, either of which construes grammatically with the rest of the 
sentence. In Rev. 2.20, f] Xeyovaa, a pendent nominative, was 
emended to rr)v Xeyovoav, which stands in apposition to the 
immediately preceding words, xi]v yvvalxa 'Ie^d/leX . 

2. Harmonistic Corruptions 

Some harmonistic alterations originated unintentionally (exam- 
ples are given on p. 257); others were made quite deliberately. Since 
monks usually knew by heart extensive portions of the Scriptures 
(see p. 127), the temptation to harmonize discordant parallels or 
quotations would be strong in proportion to the degree of the copy- 
ist's familiarity with other parts of the Bible. The words that belong 
in John 19-20, "It was written in Hebrew, in Latin, and in Greek," 
have been introduced into the text of many manuscripts at Luke 
23-38. The shorter form of the Lord's Prayer in Luke 11.2-4 ("Father, 
hallowed be your name. Your kingdom come. Give us each day our 
daily bread; and forgive us our sins, for we ourselves forgive every 
one who is indebted to us; and lead us not into temptation") was 
assimilated in many copies of Luke to agree with the more familiar, 
longer form in Matt. 6.9-13- At Acts 9-5-6, the words spoken to Paul 
at his conversion are conformed in some manuscripts to agree with 
the parallel account in 26.14-15- 

Frequently, Old Testament quotations are enlarged from the Old 
Testament context or made to conform more closely to the Septu- 
agint wording. For example, the clause in the King James Version at 
Matt. 15-8, "[This people] draweth nigh unto me with their mouth" — 
a clause that is not found in the earlier manuscripts of Matthew — was 
introduced into later manuscripts by conscientious scribes who com- 
pared the quotation with the fuller form in the Septuagint of Isa. 

Causes of Error in Transmission of the Text 263 

29.13. The earlier manuscripts of John 2.17 quote Ps. 69-9 in the form 
"Zeal for your house will consume (xara^dyerai) me." Since, how- 
ever, the current Septuagint text of this Psalm reads the aorist form 
(xarecpays), later scribes conformed the Johannine quotation to the 
text of the Septuagint. At Rom. 13.9, Paul's reference to four of the 
Ten Commandments is expanded in some manuscripts by the addi- 
tion of another "You shall not bear false witness." At Heb. 12.20, a 
few witnesses extend the quotation from Exod. 19.13 "If even a beast 
touches the mountain, it shall be stoned" by adding the words that 
follow in Exodus, "or thrust through with a dart" (as the King James 
Version renders it). 

3. Addition of Natural Complements 
and Similar Adjuncts 

The work of copyists in the amplifying and rounding off of 
phrases is apparent in many passages. Not a few scribes supposed 
that something is lacking in the statement in Matt. 9.13 "For I came 
not to call the righteous, but sinners" and added the words "unto 
repentance" (from Luke 5.32). So, too, many a copyist found it hard 
to let "the chief priests" pass without adding "the scribes" (e.g. Matt. 
26.3) or "scribes" without "Pharisees" (e.g., Matt. 27.41) or to copy 
out the phrase "Your Father who sees in secret will reward you" 
(Matt. 6.4, 6) without adding "openly." 

Col. 1.23 contains an interesting example illustrating how scribes 
succumbed to the temptation of enhancing the dignity of the apostle 
Paul. In this verse, the author warns the Colossians against shifting 
from the hope of the gospel, which "has been preached to every 
creature under heaven and of which I, Paul, became a minister." The 
word didxovog, which means literally "one who serves," a minister, 
also came to be used for a lower order of the ministry ("deacon"). 
Perhaps thinking that such a rank was less than appropriate for the 
great apostle to the Gentiles, the scribes of r* and P changed 
duxxovoq to xfjgvE, xal dnooxoXoq, while A, Syr 11 mg , and one 
manuscript of the Sahidic read all three nouns ("of which I, Paul, be- 
came a herald and apostle and minister"). MS. 81 reads bidxovoc, xal 
djTooroXog, and the Ethiopic prefers xfjgvi; xal didxovog. Here, the 
shorter, less spectacular reading is obviously original. 

A good example of a growing text is found in Gal. 6.17, where 
the earliest form of the text is that preserved in p 46 , B, A, C*, and/ "I 
bear on my body the marks of Jesus." Pious scribes could not resist 

264 The Text of the New Testament 

the temptation to embroider the simple and unadorned 'Itjoov with 
various additions, producing xvqiov 'Itjoov, as in C 3 , D c , E, K, L, and 
many other witnesses; kvqiov 'Irjoov Xqlotov in s, d, e, and Aug; 
and kvqlov rjficHv lt]oov Xqlotov in D gr *, F, G, Old Latin, Syr p , Goth, 
Chr, Vict, and Epi. 

4. Clearing Up Historical and 
Geographical Difficulties 

In the earlier manuscripts of Mark 1.2, the composite quotation 
from Malachi (3-D and from Isaiah (40.3) is introduced by the for- 
mula "As it is written in Isaiah the prophet." Later scribes, sensing 
that this involves a difficulty, replaced ev tw 'Hoata rm noo<pr]xr} 
with the general statement ev tolc, JiQOfprJTatc;. Since the quotation 
that Matthew (27.9) attributes to the prophet Jeremiah actually comes 
from Zechariah (11.12 f.), it is not surprising that some scribes 
sought to mend the error, either by substituting the correct name or 
by omitting the name altogether. A few scribes attempted to harmo- 
nize the Johannine account of the chronology of the Passion with 
that in Mark by changing "sixth hour" of John 19.14 to "third hour" 
(which appears in Mark 15.25). At John 1.28, Origen 18 altered 
Brjdaviq to Brjda^agq in order to remove what he regarded as a 
geographical difficulty, and this reading is extant today in MSS. C c , K, 
T" d , W c , fam. 13, and many others, including those that lie behind 
the King James Version. The statement in Mark 8.31 that "the Son of 
man must suffer many things . . . and be killed and after three days 
(perd Toeig rffiegag) rise again," seems to involve a chronological 
difficulty, and some copyists changed the phrase to the more famil- 
iar expression "on the third day" {xfj XQay tffiEQq). 

The author of the Epistle to the Hebrews places the golden altar 
of incense in the Holy of Holies (Heb. 9.4), which is contrary to the 
Old Testament description of the Tabernacle (Exod. 30.1-6). The 
scribes of Codex Vaticanus and of manuscripts of the Sahidic version 
correct the account by transferring the words to 9-2, where the 
furniture of the Holy Place is itemized. 

18. Comment on John, ii.19 (p. 13). See Bart D. Ehrman, Gordon D. 
Fee, and Michael W. Holmes, The Text of the Fourth Gospel in the Writings 
of Origen (Atlanta, 1992), p. 72, esp. n. 22. 

Causes of Error in Transmission of the Text 265 

5. Conflation of Readings 

What would a conscientious scribe do if the same passage was 
given differently in two or more manuscripts that were available? 
Rather than make a choice between them and copy only one of the 
two variant readings (with the attendant possibility of omitting 
the genuine reading), many scribes incorporated both readings in 
the new copy that they were transcribing. This produced what is 
called a conflation of readings and is characteristic of the later, 
Byzantine type of text. For example, in some early manuscripts, the 
Gospel according to Luke closes with the statement that the disciples 
"were continually in the temple blessing God," while others read 
"were continually in the temple praising God." Rather than discrimi- 
nate between the two, later scribes decided that it was safest to put 
the two together, so they invented the reading "were continually in 
the temple praising and blessing God." 

In the early manuscripts, at Mark 13.11, Jesus counsels his 
followers "do not worry beforehand" (jtQOjUsgi/uvdTe), concerning 
what they should say when persecuted. Other manuscripts of Mark 
read "do not prepare your defense in advance" (^QOfieXeTdre), 
which is the expression used also in the Lucan parallel (21.14). 
Rather than choose between these two versions, a good many copy- 
ists of Mark gave their readers the benefit of both. In Acts 20.28, the 
two earlier readings "church of God" and "church of the Lord" 
are conflated in later manuscripts, producing "the church of the Lord 
and God." 

Occasionally, conflate readings appear even in early manu- 
scripts. For example, Codex Vaticanus is alone in reading xakioavxi 
xai ixav(baavTi at Col. 1.12, whereas all the other manuscripts have 
one or the other participle. 

6. Alterations Made Because of 
Doctrinal Considerations 

The number of deliberate alterations made in the interest of doc- 
trine is difficult to assess. 19 Irenaeus, Clement of Alexandria, Tertul- 
lian, Eusebius, and many other Church fathers accused the heretics 
of corrupting the Scriptures in order to have support for their special 

19. See Ehrman, Orthodox Corruption of Scripture. 

266 The Text of the New Testament 

views. 20 In the mid-second century, Marcion expunged his copies of 
the Gospel according to Luke of all references to the Jewish back- 
ground of Jesus. Tatian's harmony of the Gospels contains several 
textual alterations that lent support to ascetic or encratitic views. 

Even among orthodox Christians one party often accused 
another of altering the text of the Scriptures. Ambrosiaster, the 
fourth-century Roman commentator on the Pauline Epistles, believed 
that where the Greek manuscripts differed on any important point 
from the Latin manuscripts that he was accustomed to use, the 
Greeks "with their presumptuous frivolity" had smuggled in the 
corrupt reading. In revising the Old Latin text of the Gospels, Jerome 
was apprehensive lest he be censured for making even slight alter- 
ations in the interest of accuracy — a fear that events proved to be 
well founded! 

The manuscripts of the New Testament preserve traces of two 
kinds of dogmatic alteration: those that involve the elimination or al- 
teration of what was regarded as doctrinally unacceptable or incon- 
venient and those that introduce into the Scriptures "proof for a 
favorite theological tenet or practice. 

In transcribing the prologue to the third Gospel, the scribes of 
several Old Latin manuscripts as well as the Gothic version obviously 
thought that the Evangelist should have referred to divine approval 
of his decision to compose a Gospel, so to Luke's statement (1.3), "It 
seemed good to me ... to write an orderly account" they added after 

20. See August Bludau, Die Schriftfalschungen der Haretiker. ein 
Beitrag zur Textkritik der Bibel (Miinster, 1925). Such changes prove that 
the autographs of the books of the New Testament were no longer in exis- 
tence, otherwise an appeal would have been made directly to them. Their 
early loss is not surprising, for during persecutions the toll taken by imper- 
ial edicts aiming to destroy all copies of the sacred books of Christians 
must have been heavy. Furthermore, simply the ordinary wear and tear of 
the fragile papyrus, on which at least the shorter Epistles of the New Testa- 
ment had been written (see the reference to %a.Qxr\c, in 2 John, verse 12), 
would account for their early dissolution. It is not difficult to imagine 
what would happen in the course of time to one much-handled manu- 
script, passing from reader to reader, perhaps from church to church (see 
Col. 4.16) and suffering damage from the fingers of eager, if devout readers 
as well as from climatic changes. (On Peter of Alexandria's reference to the 
original copy of John's Gospel preserved at Ephesus [Migne, Patrologia 
Graeca, 28.5171, see Juan Leal, "El aut6grafo de IV Evangelio," Estudios 
eclesidsticos, xxxiv [I960], pp. 895-905) 

Causes of Error in Transmission of the Text 267 

"me" the words "and to the Holy Spirit." The addition imitates the 
text of Acts 15.28, which reads "For it has seemed good to the Holy 
Spirit and to us. . . ." 

The inconsistency between Jesus' declaration in John 7.8 "I am 
not going to this festival, for my time has not yet fully come" and the 
statement two verses later "But after his brothers had gone up to the 
feast, then he also went up, not publicly but in private" (a discrepancy 
that Porphyry 21 seized upon to accuse Jesus of "inconstantia ac 
mutatio") led some scribes to change ovx to ovjhd ("I am not yet 
going up"). Also, Jesus' statement "But about that day and hour no 
one knows, neither the angels of heaven, nor the Son, but only the 
Father" (Matt. 24.36 and Mark 13.32) was unacceptable to scribes who 
could not reconcile Jesus' ignorance with his divinity and who saved 
the situation by simply omitting the phrase ovde 6 viol- 
in Luke 23.32, the text of p 75 , K, and B reads "Hyovxo bz xai 
eteqoi xaxovgyoi dvo ovvavtw avaigeOfjvai ("And also other crim- 
inals, two, were led away with him to be crucified"). To avoid the 
implication that Jesus was also a criminal, most Greek witnesses 
have changed the sequence of words to etsqoi dvo xaxovgyoi, 
which has the effect of subordinating the word xaxovgyoi ("And 
also two others, criminals, were led away with him to be crucified"). 
Two Old Latin manuscripts (c and e), the Sinaitic Syriac, and the 
Sahidic version solve the difficulty in another way — they leave 
Exegoi untranslated. 

An interesting variant reading, reflecting a certain delicate per- 
ception of what was deemed to be a more fitting expression, is 
found in one manuscript of the Palestinian Syriac lectionary at Matt. 
12.36; instead of the generally received logion of Jesus, "I tell you, 
on the day of judgment you will have to give an account for every 
careless word you utter," the scribe of Codex c wrote "people will 
render account for every good word they do not utter." 

In Luke 2, there are several references to Joseph and Mary 
which, in the ordinary text, doubtless appeared to some persons in 
the early Church to require rephrasing in order to safeguard the 
virgin birth of Jesus. In 2.41 and 43, instead of the words "his par- 
ents" (olyoveigavTov), some manuscripts read "Joseph and Mary." 
In 2.33 and 48, certain witnesses alter the reference to Jesus' father 

21. Quoted by Jerome, Dialogus contra Pelagianos, 2.17 (Migne, P.L. 
xxiii, pp. 578 f.). 

268 The Text of the New Testament 

either by substituting the name Joseph (as in verse 33) or by omitting 
it altogether (as in verse 48). 

In view of the increasing emphasis on asceticism in the early 
Church and the corresponding insistence upon fasting as an obligation 
laid on all Christians, it is not surprising that monks, in their work of 
transcribing manuscripts, should have introduced several references 
to fasting, particularly in connection with prayer. This has happened 
in numerous manuscripts at Mark 9.29, Acts 10.30, and 1 Cor. 7.5. In 
Rom. 14.17, where the kingdom of God is said to be not eating and 
drinking "but righteousness and peace and joy in the Holy Spirit," 
Codex 4 inserts after "righteousness" the words "and asceticism" (xai 
aoxrjoig). Such interpolations abound in 1 Corinthians 7. 22 

7. Addition of Miscellaneous Details 

In Matt. 1.8, Codex Bezae and the Curetonian Syriac insert sev- 
eral additional Old Testament names into Jesus' genealogy, thereby 
destroying the Evangelist's intended pattern of 14 generations (1.17). 
Besides the instances of agrapha contained in certain manuscripts at 
Luke 6.4 and Matt. 20.28 (see p. 71), there is a curious expansion of 
Jesus' words to Peter in a twelfth- or thirteenth-century minuscule 
codex of the Gospels (no. 713) at Matt. 17.26. The passage runs as 
follows (the addition is in italics): 

Jesus spoke of it first, saying, "What do you think, Simon? From whom 
do kings of the earth take toll or tribute? From their sons or from 
others?" When Peter said, "From others," Jesus said to him, "Then the 
children are free." Simon said, "Yes." Jesus said to him, "Then you also 
must give, as being an other to them. However, so that we do not give 
offense to them, go to the sea and cast a hook, etc." 

It is noteworthy that this expansion, preserved in a late Greek man- 
uscript, was apparently known in the second or third century, for it 
is witnessed by Ephraem's commentary on Tatian's Diatessaron as 
well as the Arabic form of the diatessaron. 23 

22. For additional variations related to social conflicts in early Chris- 
tianity, including the opposition to women, antagonism toward Jews, and 
defenses of the faith, see the discussion on pp. 287-91. 

23. For a discussion of the reading, see J. Rendel Harris, "The First 
Tatian Reading in the Greek New Testament," Expositor, 8th ser., xxiii 
(1922), pp. 120-9. 

Causes of Error in Transmission of the Text 269 

Two late minuscule manuscripts of the Book of Acts (614 and 
2147) describe the Philippian jailor as mordgErecpavag (Acts 16.27). 
Codices 181 and 460 identify the members of "the household of 
Onesiphorus," to whom the writer of 2 Tim. 4.19 sends greetings; in 
accordance with the apocryphal Acts of Paul and Thecla, they are 
said to be "Lectra, his wife, and Simaeas and Zeno, his sons." 

In the Vulgate at Phil. 4.3, the words yvr/oie ov^vye ("true yoke- 
fellow") are appropriately rendered by the Latin germane compar. 
Curiously enough, the Greek text of the bilingual manuscripts F and 
G make the adjective germane into a proper name, reading yvr\oi£ 
rsQuave ovt,vye\ 

The threefold sanctus, ayiog, ayiog, ayiog, sung by the four 
living creatures before the throne of God (Rev. 4.8), is expanded in 
various manuscripts; according to Hoskier's collations, one or more 
manuscripts have ayiog four times, six times, seven times, eight times 
(»*), nine times (B and 80 other manuscripts), and even 13 times 
(MS. 2000). 

According to a scribal addition in the margin of Codex S, the 
name of Cleopas' companion on the Emmaus road (Luke 24.18) was 
Simon (o fiExa zov KXecojta Jiooevdfievog Hi/xov r\v, ov% 6 Tlergog 
aXX 6 eregog). The margin of Codex V has the note "The one with 
Cleopas was Nathanael, as the great Epiphanios said in the 
Panarion. Cleopas was a cousin of the Savior, the second bishop of 

A number of interesting expansions appear in manuscripts of the 
early versions. An apocryphal addition in two Old Latin manuscripts 
(a and g 1 ) states that when Jesus "was baptized, a tremendous light 
flashed forth from the water, so that all who were present feared" 
(Matt. 3.15). Another Old Latin manuscript (k) amplifies Mark's 
account of the resurrection of Jesus by adding at 16.3: 

Suddenly at the third hour of the day there was darkness throughout 
the whole circuit of the land, and angels descended from heaven, and 
he rose in the brightness of the living God, [and] at once they ascended 
with him, and immediately there was light. Then they [the women] 
drew near to the tomb. 24 

The natural curiosity of readers regarding the identity of persons 
who are referred to without being named in the New Testament 

24. The Latin text at "Third hour of the day" and at "he rose" is not 
grammatical; the English renders the general sense. 


The Text oe the New Testament 

prompted scribes to supply proper names. 25 The Sahidic version 
gives the name Nineveh to the anonymous rich man of Luke 16.19, 
as does the Bodmer papyrus p 75 (see p. 59). The two robbers who 
were crucified on either side of Jesus are variously named in Old 
Latin manuscripts: 26 



Codex c 

Matt. 27.38 



Mark 15.27 





Codex / 

Luke 23.32 



Codex r 

Luke 23.32 



The titles of the books of the New Testament were the objects of 
a good deal of elaboration by scribes. It will be obvious that titles of, 
for example, Paul's Epistles were not needed until the apostle's cor- 
respondence had been collected into one corpus. The earliest titles 
were short and to the point. Later scribes, however, were not content 
with a bare and unadorned title; they embroidered it in ways they 
thought to be in accord with the position and reputation of the au- 
thor. Thus, in » and C, the Book of Revelation is entitled simply 
'AnoxdXvipic, 'Icodvvov. Later manuscripts describe John as "the di- 
vine" (i.e., "the theologian," 'AnoxdXvipig 'Icodvvov tov deoXoyov, 
MSS. 35, 69, 498, and 1957). Others expand by prefixing "saint" to the 
name (dyiov 'Icodvvov; MSS. 1, 2015, 2020, etc.), and still others add 
"the Evangelist" and/or "the apostle." The longest and most fulsome 
title is that found in a manuscript at Mount Athos (Hoskier's 236; 
Greg. 1775): "The Revelation of the all-glorious Evangelist, bosom 
friend [of Jesus], virgin, beloved to Christ, John the theologian, son 
of Salome and Zebedee, but adopted son of Mary the Mother of 
God, and Son of Thunder" CH djtoxdXvipig tov navevdo^ov 
evayyekiaxov, imorrjdiov (f>iXov, Jtagdsvov, fiyamjuevov too 
Xqlotoj, 'Icodvvov tov OeoXoyov, vlov HaXcofirjg xai Zefiedaiov, 

25. See Bruce M. Metzger, "Names for the Nameless in the New 
Testament: A Study in the Growth in Christian Tradition," in Kyriakon: 
Festchrift Johannes Quasten, ed. by Patrick Granfield and Josef A. 
Jungmann (Munster, 1970), pp. 79-99. 

26. See J. Rendel Harris, "On Certain Obscure Names in the New 
Testament," Expositor, 6th ser., i (1900), pp. 161-77; idem, "A Further Note 
on the Names of the Two Robbers in the Gospel," Expositor, 6th ser., 

i (1900), pp. 304-8. 

Causes of Hrror in Transmission of the Text 271 

Qetov de viov xfjg deoxoxov Magiag, xai viov figovrifg). The only 
designation that the scribe omits (probably by accident!) is "apostle." 

Other scribal additions that eventually found their way into the 
King James Version are the subscriptions appended to the Pauline 
Epistles, giving information regarding the traditional place from 
which each was sent as well as in some cases what was believed to 
be the name of the amanuensis or of the messenger who was to 
carry the epistle. 

Lest the foregoing examples of alterations should give the im- 
pression that scribes were altogether willful and capricious in trans- 
mitting ancient copies of the New Testament, it ought to be noted 
that other evidence points to the careful and painstaking work on 
the part of many faithful copyists. There are, for example, instances 
of difficult readings that have been transmitted with scrupulous 
fidelity. Thus, f/Xdev at Gal. 2.12 yields no good sense and can 
scarcely be the form intended by the author. Nevertheless, the 
scribes of the earliest manuscripts (including p /|6 , K, B, D*, and G) 
refrained from correcting it to r\XBov. Another instance of a mani- 
festly erroneous reading is el'ng anXdy/ya xai olxTiQfioi at Phil. 2.1, 
which could have arisen when the original amanuensis misunder- 
stood Paul's pronunciation of eX xi cmXdyyya. . . . However the sole- 
cism may have originated, the point is that all majuscules and most 
minuscules have transmitted it with conscientious exactness. 

Even in incidental details one observes the faithfulness of 
scribes. For example, the scribe of Codex Vaticanus copied quite 
mechanically the section numbers that run in one series throughout 
the corpus of the Pauline Epistles, even though this series had been 
drawn up when the Epistle to the Hebrews stood between Galatians 
and Ephesians and is therefore not suitable for the present sequence 
of the Epistles in Vaticanus. 27 These examples of dogged fidelity on 
the part of scribes could be multiplied and serve to counterbalance, 
to some extent, the impression that this chapter may otherwise make 
upon the beginner in New Testament textual criticism. 28 

27. See p. 69, n. 27. 

28. For other examples of the faithfulness of scribes, see B. 
Blumenkranz, "Fidelite du scribe," Revue du Moyen Age, viii (1952), 

pp. 323-6; H. J. Vogels, "Xibrarii dormitantes': Aus der Uberlieferung des 
Ambrosiaster-Kommentars zu den Paulinischen Briefen," Sacris Erudiri: 
Jaarboek voor Codsdienstwetenscbappen, viii (1956), pp. 5-13. 


History of the Transmission of 
the Text of the New Testament 

For the textual critic interested in establishing the original text of the 
New Testament, it is important to know about the history of textual 
transmission, from the earliest of times down through the Middle 
Ages. Among other things, this means knowing about the rise and 
development of the major textual groupings — that is, the various text 
types represented among the manuscripts. Even though many of the 
details are shrouded in mystery, it is possible to trace the general 
outline of this history and to evaluate the textual character of the 
major text types attested among our surviving witnesses. 

I. Complications in Establishing 
the Original Text 

To some extent, each of the books of the New Testament has its 
own textual history. What can be said of all the books, however, is that 
each first appeared as a discrete publication, or series of publications, 1 
by its author(s). It is relatively easy to pinpoint the origin of some of 
these texts: Paul's letter to the Galatians, for example, was written at a 

1. It is possible, for example, that the Gospel of John was published 
in several editions; one of the earlier editions may well have concluded at 
20.31, with chapter 21 being added at a later time. See the introduction 


History of the Transmission of the Text 273 

certain time and place to a certain audience. But other texts present 
complications.- many scholars have considered 2 Corinthians, for 
example, to be a combination of two or more Pauline letters, each 
written at a different time, for a different occasion, and only later com- 
bined into the one letter we now have. Moreover, there is at least one 
passage in 2 Corinthians that does not appear to have been part of the 
original letter(s) at all but to have been interpolated into the letter at a 
later time by a later hand (6. 14—7. 1). This may be true of other passages 
in the Pauline corpus as well; scholars today continue to debate the 
possibility of non-Pauline interpolations that occurred before any of 
the surviving manuscripts had been produced. 2 

These examples show why it is difficult — some would say 
impossible — to talk about the original text of the Pauline epistles. For 
example, which form of 2 Corinthians is the original? Is it the text of 
the separate letters that were eventually combined into the one let- 
ter we now have? It would seem to be an impossible task to estab- 
lish those texts since they now exist only in their edited form, with 
parts omitted and/or edited to form the one letter of 2 Corinthians. 
Is it the first edition of this one conglomerate letter that should be re- 
constructed? If so, should that letter be reconstructed with or without 
the interpolation in 6.14-7.1? Moreover, since Paul evidently dictated 
his letters, what if the amanuensis who recorded his words made a 
mistake? Is the original text the one with the mistake? Or is it the text 
that Paul spoke — or meant to speak? If the latter, how can we possi- 
bly get back to an oral dictation that was erroneously recorded? 

If anything, the situation is even more complicated with the 
Gospels, for these are based on oral traditions and written sources to 
which we no longer have independent access. What would it mean to 
reconstruct the original version of, say, the Gospel of John? Would the 
textual critic reconstruct the earliest version that excludes chapter 21? 

in Raymond Brown, The Gospel According to John (Garden City, NY, 1966). 
Likewise, scholars from Blass at the end of the nineteenth century to 
Boismard and Lamouille at the end of the twentieth have maintained that 
the Book of Acts was published in two editions, one that is now repre- 
sented in the Alexandrian tradition and the other in the Western, which is 
some 8.5% longer than the other. See M.-E. Boismard and A. Lamouille, 
Le texte occidental des Actes des Apotres; reconstruction et rehabilitation: 
vol. i, Introduction et texts; vol. ii, Apparat critique (Paris, 1984). 

2. See William O. Walker, Interpolations in the Pauline Letters 
(London, 2001). 

274 Thf. Text of the New Testament 

Probably not, since all of our (complete) Greek manuscripts contain 
that chapter. Then, would it mean reconstructing the final edition of 
the Gospel, which includes the chapter? If so, should we consider the 
story of the woman taken in adultery (7.53-8.11) to be part of the final 
edition or accept that this account was added too late to be consid- 
ered original? 3 

These are some of the theoretical issues that the textual critic must 
face and resolve in dealing with the most basic of questions: what 
does it mean to establish an "original" text? 4 The complications 
notwithstanding, most textual critics would agree that our manuscripts 
for each biblical book, whether Galatians, 2 Corinthians, or John, in 
one way or another go back to a text that was produced — either au- 
thored or edited — and published at some specific time and place and 
that it is this "final published" edition that served as the basis for all 
later copies that the textual critic is trying to reconstruct. 5 

II. Dissemination of Early 
Christian Literature 

Once each of the books of the New Testament was brought to 
completion, it was put into circulation among the Christian communi- 
ties. This is what it meant to publish a book in antiquity: to provide one 
or more copies (each made separately, by hand) for others to read. 6 

3. For further reflections on such problems in reference to the 
Gospels, see William Petersen, "The Genesis of the Gospels," in New 
Testament Textual Criticism and Exegesis. Festschrift J. Delobel, ed. by 

A. Denaux (Leuven, 2002), pp. 33-65. For a fuller discussion, see David C. 
Parker, The Living Text of the Gospels (Cambridge, 1997). 

4. See Eldon Jay Epp, "The Multivalence of the Term 'Original Text' in 
New Testament Textual Criticism," Harvard Theological Review, xcii (1999), 
pp. 245-81. 

5. A further complication presents itself in the case of the letters of 
Paul, in that many scholars have come to think that all of our surviving 
manuscripts derive ultimately not from the "originals" that Paul produced 
but from a collection of Paul's writings that was made sometime near the 
end of the first century. If that is the case, it would be difficult to get 
behind the texts presented in that collection to the original texts produced 
some 40 years earlier. See especially Gunther Zuntz, The Text of the Epistles: 
A Disquisition upon the Corpus Paulinum (London, 1953). 

6. See Harry Y. Gamble, Books and Readers in the Early Church (New 
Haven, 1995), pp. 42-53. 

I Iistory of the Transmission of the Text 275 

The earliest copies of each of the books of the New Testament 
would no doubt have been made either in the community in which 
the book was first produced (e.g., if Paul made an extra copy of a 
letter before sending it off) or in the community to which it was ad- 
dressed. As other Christians wanted additional copies either for them- 
selves or for their communities, these too would need to be made by 
hand. The earliest copyists would not have been trained professionals 
who made copies for a living but simply literate members of a con- 
gregation who had the time and ability to do the job. Since most, if not 
all, of them would have been amateurs in the art of copying, a rela- 
tively large number of mistakes no doubt crept into their texts as they 
reproduced them. It is possible that after the original was placed 
in circulation it soon became lost or was destroyed, so all surviving 
copies may conceivably have derived from some single, error-prone 
copy made in the early stages of the book's circulation. 

As the Christian Church spread throughout the Mediterranean 
world, with new congregations springing up in major urban areas, 
the literature of the early Christians proliferated and such amateur 
copies multiplied. By the early second century, virtually all of the 
books that were eventually included in the New Testament had 
been produced — written, edited, and circulated — and multiple 
copies would have already been available in different locales. 
Within each locale, texts would have been copied and recopied as 
demand would have grown with the rise of additional churches and 
the conversion of literate persons to the new faith, some of whom 
would have wanted copies of apostolic writings for themselves or 
for their communities of faith. Conceivably, different localities would 
have had each of these texts in some distinctive form, as scribes 
within the community copied the community's copies, which may 
have differed in greater or lesser ways from the copies of other 

We have good evidence to indicate that in the early decades of 
transmission numerous changes were made to the texts in circula- 
tion: as words or entire lines came to be left out inadvertently or 
inadvertently copied twice, stylistic changes were made, words were 
substituted for one another, evident infelicities or outright mistakes 
were corrected, and so on (see the discussion of the kinds of change 
in Chapter 7). It is a striking feature of our textual record that the 
earliest copies we have of the various books that became the New 
Testament vary from one another far more widely than do the later 
copies, which were made under more controlled circumstances in 
the Middle Ages. Moreover, the quotations of the New Testament by 

276 The Text of the New Testament 

early church fathers evidence a wide array of textual variation dating 
from these earliest stages in the history of transmission. 

III. The Rise and Development of the New 
Testament Text Types 

By the middle of the second century, there were sufficient num- 
bers of manuscripts available that they could begin influencing one 
another, to the extent that a copyist was able at specific passages to 
compare one exemplar with another to decide which form of the 
text to copy. On the one hand, this kind of mixture complicates the 
textual picture, making it almost impossible to trace the direct 
genealogical descent of manuscripts (since the textual "lines of 
descent" are mixed and overlapping). On the other hand, it also 
means that certain kinds of change — for example, harmonizations, 
loose paraphrases, substitution of a more common word for a less 
common one — came to be made more consistently among the man- 
uscripts so that copies not only embodied similar kinds of change 
but actually reproduced the very same changes. 

1 . The Western Text 

Most scholars have come to see that the Western text is a prod- 
uct of these second-century copying practices. 7 The Western text is 
not aptly named, as it has been found as well outside the western 
parts of the Roman Empire. Moreover, some scholars have called 
into question its existence as a text type because the witnesses that 
attest it do not do so with strict consistency, lacking the kind of 
homogeneity one finds in other types of text. Nonetheless, there are 
both certain kinds of change reflected in Western witnesses and (to 
a lesser extent than with other text types) specific readings shared 
among them. Because of the loose association of the various 
witnesses of the Western text, most scholars do not consider it the 
creation of an individual or several individuals revising an earlier 
text but, rather, the result of the undisciplined and "wild" growth of 
manuscript tradition in the second century. 

7. For a recent overview of scholarship, see Joel Delobel, "Focus on 
the 'Western' Text in Recent Studies," Ephemerides Theologicae Lovanienses, 
lxxiii (1997), pp. 401-10. 

History of the Transmission of the Text 277 

The Western type of text can be traced to a very early elate, for it 
was used by Marcion, Justin, Heracleon, Irenaeus, Tertullian, and 
other patristic sources of the second century. 8 Among the papyri it 
can be found in p 48 (about the end of the third century) and p 38 
(around A.D. 300). Its most important witnesses are Codex Bezae and 
the Old Latin manuscripts, all of which are characterized by longer or 
shorter additions and by certain striking omissions. So-called Western 
texts of the Gospels, Acts, and Pauline Epistles circulated widely, 9 not 
only in North Africa, Italy, and Gaul (which are geographically west- 
ern) but also in Egypt 10 and (in somewhat different text forms) the 
East. These latter text forms are represented by the Sinaitic and Cure- 
tonian manuscripts of the Old Syriac, by many of the marginal notes 
in the Harclean Syriac, and perhaps by the Palestinian Syriac. 

The chief characteristic of Western readings is fondness of para- 
phrase. Words, clauses, and even whole sentences are freely changed, 
omitted, or inserted. Sometimes the motive appears to have been 
harmonization, while at other times it was the enrichment of the 
narrative by inclusion of traditional or apocryphal material. 

2. The Alexandrian Text 

It would be a mistake to think that the uncontrolled copying 
practices that led to the formation of the Western textual tradition 
were followed everywhere that texts were reproduced in the Roman 
Empire. In particular, there is solid evidence that in at least one major 
see of early Christendom, the city of Alexandria, there was conscious 

8. See, e.g., Bart D. Ehrman, "Heracleon and the 'Western' Textual 
Tradition," New Testament Studies, xl (1994), pp. 465-86. The patristic at- 
testation of the Western text shows that K. Aland was not right to maintain 
that it came into existence only during the second half of the third century, 
when the church was free from persecution (from A.D. 260 to 303); 
Aland's view was based on his analysis of the papyri and other manuscript 
witnesses. See his "Alter und Entstehung des D-Textes im Neuen Testa- 
ment: Betrachtungen zu p 69 und 0171," Miscelania papirologica Ramon 
Roca-Puig, ed. by Sebastia Janeras (Barcelona, 1987), pp. 37-61, esp. p. 43. 

9. The General Epistles and the Book of Revelation seem not to have 
existed in a characteristically Western form of text. 

10. Evidence for the presence of the Western text in Egypt is found 
chiefly in several papyri (e.g., p 29 , p 38 , and p' 48 ). The common opinion that 
Clement of Alexandria was accustomed to use a Western form of text 
(based on P. M. Barnard, The Biblical Text of Clement of Alexandria [Texts 

278 Thk Text of the New Testament 

and conscientious control exercised in the copying of the books of 
the New Testament. 

Alexandria, widely known throughout the ancient word as a 
major center for learning and culture, had a long history of classical 
scholarship, attached principally to its famous museum and library 
but influential on larger parts of its population. 11 It is no surprise, 
then, to find that textual witnesses connected to Alexandria attest a 
high quality of textual transmission from the earliest times. It was 
there that a very ancient line of text was copied and preserved, as is 
evidenced in such Alexandrian church writers of the third and fourth 
centuries as Origen, Athanasius, and Didymus the Blind, in such no- 
table manuscripts as p 66 , p 7 ^, Codex Vaticanus, and Codex Sinaiticus, 
and in copies of the Coptic versions. In light of the striking similari- 
ties in text between the fourth-century B and the early third-century 
p 7 \ it is clear that the Christian scholars of Alexandria worked assid- 
uously to preserve an accurate form of text. 

Even in Alexandria, of course, changes were made to the texts 
that were copied and recopied; for that reason, Alexandrian witnesses 
today are classified according to whether they preserve the excellent 
p /5 -B line of text (these are the primary Alexandrian witnesses, which 
would include p 66 [c. 200] and Origen) or instead attest alterations of 
this line of text that sometimes involve grammatical and stylistic pol- 
ishing (these are the secondary Alexandrians, including manuscripts 
C, L, 33, and the writings of Didymus the Blind). 

In the second and third centuries, then, we have evidence of two 
major text types, the paraphrastically inclined and somewhat uncon- 
trolled Western text and the carefully preserved and relatively pris- 
tine Alexandrian text. No doubt other distinctive textual traditions 
were formed and developed during this period as well; unfortu- 
nately, few traces of such local texts remain today. 12 

and Studies, v. 5; Cambridge, 18991) must be modified in the light of 
further research by R. J. Swanson ("The Gospel Text of Clement of 
Alexandria" [Diss., New Haven, 19561), who found that in the Stromateis 
Clement's quotations of Matthew and John are twice as often from the 
Egyptian (Alexandrian) text as from the Western text. See further Michael 
Mees, Die Zitate aus dem Neuen Testament bei Clemens von Alexandrien 
(Rome, 1970). 

11. See Rudolf Pfeiffer, History of Classical Scholarsh ip: From the Begin- 
nings to the End oj the Hellenistic Age (Oxford, 1968), pp. 87-104, 171-209. 

12. Including the elusive Caesarean text. See pp. 310-2. 

History of the Transmission of the Text 279 

3. The Byzantine Text 

The other major textual tradition to survive is the Byzantine text, 
sometimes also known as the Syrian text (so Westcott and Hort), the 
Koine text (so von Soden), the Eclessiastical text (so Lake), and the 
Antiochian text (so Ropes). With the exception of those scholars who 
continue to appeal to the "majority text" in making textual deci- 
sions, 13 nearly all critics today see the Byzantine text as a later de- 
velopment in the history of transmission. Recent studies of the 
Byzantine text have shown that it can be found in rudimentary form 
as early as the fourth century in such church writers as Basil the 
Great and Chrysostom 14 but that its final form represents a slowly de- 
veloping tradition, not one that sprang up immediately at one time 
and place. 1 "' It was not, in other words, a textual recension created 
by a single person or community. It does appear, however, that the 
Byzantine editors formed their text by taking over elements of the 
earlier extant traditions, choosing variant readings from among those 
already available rather than creating new ones that fit their sense of 
an improved text. 16 

The Byzantine text is characterized by lucidity and complete- 
ness. Those who framed this text over a long period of time sought 
to smooth away any harshness of language, to combine two or more 
divergent readings into one expanded reading (called conflation), 

13. See pp. 218-22. 

14. See Jean-Francois Racine, Ttoe Text of Matthew in the Writings of 
Basil ofCaesarea (Atlanta, 2004). 

15. See Klaus Wachtcl, Der byzantinische Text der katholischen Briefe: 
eine Untersuchung zur Entstebung der Koine des Nenen Testaments (Berlin, 
1995). For a brief but characteristically compelling earlier statement, see 

E. C. Colwell, "Method in Establishing the Nature of Text-Types of New 
Testament Manuscripts," in Studies in Methodology in Textual Criticism of 
the New Testament (New Testament Tools and Studies; Grand Rapids, MI, 
1969), pp. 45-55. 

16. This is why newly discovered papyri will occasionally attest read- 
ings previously known only from Byzantine witnesses: the Byzantine text 
took over older readings, and in some places it provides the only access to 
them. (See H. Sturz, 'Ihe Byzantine Text-type and the New Testament 
[Nashville, 1994] and, especially, Zuntz, op. cit., p. 55) This does not mean, 
however, that the Byzantine text itself is as old as those readings, inas- 
much as the text type is a large collocation of readings in combination — in 
Colwell's words: "the entire complex of readings in its total pattern" (op. 
cit., p. 52). 

280 The Text of the New Testament 

and to harmonize divergent parallel passages. These positive charac- 
teristics are no doubt what made its readings so popular that by the 
early Middle Ages it was the text of choice among most copyists. Its 
earliest manuscript witness is the fifth-century Codex Alexandrinus 
(in the Gospels but not in Acts, the Epistles, or Revelation); it can be 
found in many of the later majuscule manuscripts and in the great 
mass of minuscule manuscripts. 

The influence of the Byzantine text was aided by historical fac- 
tors: this was the text that became popular in Constantinople, whence 
it was distributed widely throughout the Byzantine Empire (where the 
Greek language was preserved). It is no surprise, then, that it was 
the text that came to dominate the Greek textual tradition from the 
seventh century onward so that the vast majority of witnesses surviv- 
ing today are of this type. As we have seen, however, the fact that the 
bulk of witnesses attest the Byzantine text is no sign of its superiority 
when it comes to establishing the original text. To that end, the earlier 
attested text forms, the Western and most especially the Alexandrian, 
are today considered by most critics to be far superior. 

IV. The Use of Textual Data for the Social 
History of Early Christianity 17 

If the ultimate goal of textual criticism is to reconstruct the orig- 
inal text of the New Testament, criticism can be seen as a principally 
negative endeavor, involving the elimination of false readings from 
the manuscripts so as to attain the pristine originals. In the words of 
Fenton John Anthony Hort, the task involves "nothing more than the 
detection and rejection of error." 18 

17. Much of this discussion is drawn from Bart D. Ehrman, "The Text 
as Window: Manuscripts and the Social History of Early Christianity," in 
The New Testament in Contemporary Research: Essays on the Status Quaes- 
tionis, ed. by Bart D. Ehrman and Michael W. Holmes (Grand Rapids, MI, 
1995), pp. 361-79; idem, "The Text of the Gospels at the End of the Second 
Century," in Codex Bezae: Studies from the I.unel Colloquium, June 1994, 
ed. by Christian-Bernard Amphoux and David C. Parker (Leiden, 1996), 
pp. 95-122. 

18. B. F. Westcott and F. J. A. Hort, The New Testament in the Original 
Greek, ii (London, 1881; 2nd ed., 1896; repr. Peabody, MA, 1988), p. 3. 

History of the Transmission of the Text 281 

In recent decades, however, many textual critics have come to 
recognize that an exclusive concentration on the autographs can 
prove to be myopic, as it overlooks the value of variant forms of the 
text for historians interested in matters other than exegesis. Thus, 
one of the significant breakthroughs of textual scholarship has been 
the recognition that the history of a text's transmission can contribute 
to the history of its interpretation: early Christian exegetes occasion- 
ally disagreed on the interpretation of a passage because they knew 
the text in different forms. 19 

Moreover, some critics have come to recognize that variants in 
the textual tradition provide data for the social history of early Chris- 
tianity, especially during the first three Christian centuries, when the 
majority of all textual corruptions were generated. 20 This is because 
changes that scribes made in their texts sometimes reflect the socio- 
historical contexts within which they worked. By examining these 
changes, it is possible to reconstruct the contexts in which they were 
generated, contexts that are otherwise but sparsely attested in our 
surviving sources. When viewed in this way, variant readings are not 
merely chaff to be discarded en route to the original text, as they 
were for Hort; they are instead valuable evidence for the history of 
the early Christian movement. The New Testament manuscripts can 
thus serve as a window into the social world of early Christianity. 

Here, mention may be made of six areas in which the textual 
data can provide us with information concerning the social history of 
early Christianity. 

19- See, e.g., Bart D. Ehrman, "Heracleon, Origen, and the Text of the 
Fourth Gospel," Vigiliae Christianae, xlvii (1993), pp. 105-18. For a 
methodological discussion of this matter, see idem, "The Text of Mark in 
the Hands of the Orthodox," in Biblical Hermeneutics in Historical Per- 
spective, ed. by Mark Burrows and Paul Rorem (Philadelphia, 1991), 
pp. 19-31. 

20. See Bart D. Ehrman, The Orthodox Corruption of Scripture: The 
Effect of Christological Controversies on the Text of the New Testament 
(New York, 1993), pp. 28, 44, n. 112. Among earlier textual critics who 
recognized this phenomenon were F. C. Conybeare, Adolf von Harnack, 
J. Rendel Harris, Kirsopp Lake, Donald Riddle, C. S. C. Williams, and Eric 
Fascher; see ibid, p. 42, n. 94. More recent works include Eldon Epp's 
groundbreaking study The Theological Tendency of Codex Bezae Cantab- 
rigiensis in Acts (Cambridge, 1966) and David C. Parker, The Living Text 
of the Gospels (Cambridge, 1997). 

282 The Text of the New Testament 

1. Doctrinal Disputes of Early Christianity 

A pioneering study of early Christianity was Walter Bauer's 1934 
classic RechtglaiXbigkeit und Ketzerei im altesten Cbristentum. 21 
Bauer argued that orthodoxy was not an original and universally 
dominant form of Christianity in the second and third centuries, with 
heresy (in its multiple configurations) a distant and derivative sec- 
ond. Instead, early Christianity comprised a number of competing 
forms of belief and practice, one of which eventually attained domi- 
nance for a variety of social, economic, and political reasons. The 
victorious orthodoxy then rewrote the history of the church in light 
of its final triumph. This orthodoxy was the form of Christianity 
embraced by the faithful in Rome. 

While many of the details remain in serious dispute, 22 Bauer's 
overarching conception continues to exert a wide influence, as does 
his insistence on the centrality of these ideological disputes to the 
early history of Christianity. 23 Even so, many critics of the past 
century have argued that these disputes had almost no effect on the 
textual tradition of the New Testament. In part, this view was based 
on the authoritative pronouncement of Hort: "It will not be out of 
place to add here a distinct expression of our belief that even among 
the numerous unquestionably spurious readings of the New Testa- 
ment there are no signs of deliberate falsification of the text for dog- 
matic purposes." 24 Consonant with this perception was A. Bludau's 
detailed study of the charge leveled against Christian heretics of 

21. Beitrage zur historischen Theologie, 10 (Tubingen, 1934). English 
trans, of 2nd ed. (1964, ed. by Georg Strecker) by Robert Kraft et al. 
(Orthodoxy and Heresy in Earliest Christianity, ed. by Robert Kraft and 
Gerhard Krodel; Philadelphia, 1971). 

22. See, e.g., Arland J. Hultgren, The Rise of Normative Christianity 
(Minneapolis, 1994). 

23. For a useful discussion of its initial reception, see Georg 
Strecker's essay "Die Aufnahme des Buches" on pp. 288-306 of the 2nd 
German ed., expanded and revised by Robert Kraft as "The Reception of 
the Book," Appendix 2, pp. 286-316 in the English ed. The discussion was 
updated by Daniel Harrington, "The Reception of Walter Bauer's Ortho- 
doxy and Heresy in Earliest Christianity During the Last Decade," Harvard 
Theological Review, lxxiii (1980), pp. 289-98. For additional bibliography, 
see the discussion in Ban D. Ehrman, The Orthodox Corruption of Scripture, 
p. 33, n. 16. 

24. Westcott and Hort, op. cit., p. 282. Hort specifies Marcion as the 
one exception to this rule and goes on to say that non-Marcionite instances 

I listory of the Transmission of the Text 283 

intentionally falsifying the texts of Scripture, a charge that he traced 
from apostolic times to the Monophysite controversy. 25 Bludau 
argued that, in many instances, the accusation was directed not 
against heretical alterations of the text but heretical misinterpreta- 
tions; moreover, he maintained, in most of the remaining instances, 
the charges cannot be sustained. He concluded that the manuscripts 
of the New Testament were not easily susceptible to deliberate falsi- 
fication, given the vigilance exercised over their production by all 
concerned parties. 26 

Despite its popularity, this view has never held universal sway. 
Even before the Second World War, individual scholars had isolated 
and discussed instances of theologically motivated corruption, in- 
cluding such eminent figures as Kirsopp Lake, J. Rendell Harris, 
Adolf von Harnack, Donald Riddle, and most extensively, Walter 
Bauer himself (in another, less-read but equally impressive mono- 
graph). 27 Nonetheless, only within the past 40 years have scholars 
begun to recognize the full extent to which early ideological conflicts 
affected the New Testament text. A major impetus was provided by 
Eldon Jay Epp's groundbreaking study The Theological Tendency of 
Codex Bezae Cantabrigiensis in Acts, 28 whose particular conclusions 
relate more to Jewish-Christian relations (see pp. 287-8) than to the 
internecine conflicts of the early Christian movement. Nonetheless, 

of variation that appear to be doctrinally motivated are due to scribal care- 
lessness or laxity, not to malicious intent. 

25. Die Schriftfdlschungen der Hdretiker. Bin Beitrag zur Textkritik 
der Bihel (Mun&ter, 1925). 

26. For an assessment, see Ehrman, The Orthodox Corruption of Scrip- 
ture, p. 43, n. 100. 

27. See, e.g., Kirsopp Lake, The Influence of Textual Criticism on the 
Exegesis of the New Testament (Oxford, 1904); J. Rendel Harris, "New Points 
of View in Textual Criticism," Expositor, 8th ser., vii (1914), pp. 316-34; idem, 
"Was the Diatesseron Anti-Judaic?" Harvard Theological Review, xviii (1925), 
pp. 103-9; Adolf von Harnack, "Zur Textkritik und Christologie der Schriften 
Johannes," in Studien zur Geschichte des Neuen Testaments und der alten 
Kirche, vol. I, Zur neutestamentlichen Textkritik (Berlin, 193D pp. 115-27; 
idem, "Zwei alte dogmatische Korrekturen im Hebraerbrief," in Studien zur 
Geschichte des Neuen Testaments, i, pp. 235-52; Donald Wayne Riddle, 
"Textual Criticism as a Historical Discipline," Anglican Theological Review, 
xviii (1936), pp. 220-33; Walter Bauer, Das Lebenjesu im Zeitalter der 
neutestamentlichen Apocryphen (Tubingen, 1907; repr. Darmstadt, 1967). 

28. Cambridge, 1966. For Epp's predecessors, see his discussion on 
pp. 12-26. 

284 The Text of the New Testament 

Epp attacked the Hortian view head-on by pursuing the suggestion 
that some of the tendencies of the so-called Western text, as embed- 
ded in Codex Bezae, should be explained by the theological pro- 
clivities of its scribe. 29 Through a detailed and exhaustive analysis, 
Epp concluded that some 40% of Codex Bezae's variant readings in 
Acts point toward an anti-Judaic bias. The sensible inference is that 
the scribe himself, or his tradition, was anti-Jewish (in some sense) 
and that this prejudice came to be embodied in the transcription of 
the text. 30 

Subsequent analyses of theological tendencies have moved from 
the study of a specific manuscript to a panoramic view of the surviv- 
ing witnesses. 31 While no one would claim that theological contro- 
versies caused the majority of our hundreds of thousands of textual 
variants, they clearly engendered several hundred. Nor are these 
variant readings, taken as a whole, of little consequence. On the con- 
trary, many prove to be critical for questions relating to New Testa- 
ment exegesis and theology. 32 

29. A suggestion made earlier, for example, by P. H. Menoud, "The 
Western Text and the Theology of Acts," Studiorum Novi Testamenti Soci- 
etas, II (1951), pp. 27-8. 

30. A conclusion that Epp himself does not draw, as is pointed out 
on p. 288. 

31. Alexander Globe, "Some Doctrinal Variants in .Matthew 1 and Luke 2 
and the Authority of the Neutral Text," Catholic Biblical Quarterly, xlii 
(1980), pp. 52-72; Bait D. Ehrman and Mark A. Plunkett, "The Angel and the 
Agony: The Textual Problem of Luke 22:43-44," Catholic Biblical Quarterly, 
xlv (1983), pp. 401-16; Mikeal Parsons, "A Christological Tendency in p 75 ," 
Journal of Biblical Literature, cv (1986), pp. 463-79; Peter M. Head, "Chris- 
tology and Textual Transmission: Reverential Alterations in the Synoptic 
Gospels," Novum Testamentum, xxxv (1993), pp. 105-29; Ban D. Ehrman, 
The Orthodox Corruption of Scripture. 

32. The interpretation of significant passages is sometimes affected by 
the textual decision. Just within the Gospels, reference can be made to the 
Prologue of John (e.g., 1.18), the birth narratives of Matthew and Luke (e.g., 
Matt. 1.16, 18; Luke 1.35), the baptism accounts (e.g., Mark 1.10; Luke 3.22; 
John 1.34), and the various passion narratives (e.g., Mark 15.34; Luke 
22.43-4; John 19.36). Moreover, a number of variants affect a range of issues 
that continue to interest historians and exegetes of the New Testament, in- 
cluding such questions as whether the Gospels could have been used to 
support either an "adoptionistic" Christology (e.g., Mark 1.1; Luke 3.22; John 
1,34) or one that was "antidocetic" (e.g., the Western noninterpolations), 

History of the Transmission of the Text 285 

An obvious example of changes motivated by theological con- 
siderations can be found in the account of Jesus as a child in Luke 2, 
where on two occasions the oldest form of the text refers to Joseph 
as Jesus' "father" — a problematic reference for scribes committed to 
the idea (also found in Luke) that Jesus was in fact born of a virgin. 
In Luke 2.33, we are told that Jesus' "father and mother" began to 
marvel at the things being said about him. Most Greek manuscripts, 
however, along with Old Latin, Syriac, and Coptic witnesses, change 
the text to remove the problem, reading "Joseph and his mother 
began to marvel." So, too, several verses later in 2.48, Jesus' mother 
upbraids him for staying behind in Jerusalem when the rest of the 
family had started to return from the festival, by saying "Your father 
and I have been searching for you in great anxiety." This time the 
text is changed in far fewer witnesses: just one Greek manuscript of 
the fifth century and two old Latin manuscripts read "Your relatives 
and I have been grieved;" while one Syriac and several Old Latin 
manuscripts read "We have been grieved." 

In these witnesses, Jesus is no longer said to be the son of 
Joseph. This kind of change makes sense in the context of early 
Christological controversies, where some Jewish-Christian groups 
and other Christians holding to "adoptionist" Christological views 
were claiming that Jesus was a full flesh-and-blood human, the son 
of Joseph and Mary. 33 

The same polemical context may lie behind the change made 
two chapters later in Luke's Gospel, in the account of Jesus' baptism. 
In the majority of witnesses, the voice that comes from heaven 
alludes to Isaiah 42: "You are my Son, the Beloved, with you I am 
well pleased" (Luke 3-22). But in a wide range of early patristic 
sources of the second and third centuries, the voice is said to have 
quoted Ps. 2.7: "You are my Son, today I have begotten you." This 
latter form of the text, of course, could have proved useful to those 
holding to adoptionistic views, for it could be construed to say that 

whether Luke has a doctrine of the atonement (e.g., Luke 22.19-20), 
whether members of the Johannine community embraced a gnostic Christol- 
ogy (e.g., 1 John 4.3), and whether any of the authors of the New Testament 
characterizes Jesus as God (e.g., Heb. 1.8). See Ehrman, Orthodox Corrup- 
tion of Scripture, pp. 276-7. 

33. For further discussion of these variants and reflections on their 
uneven distribution among the surviving witnesses, see Ehrman, Orthodox 
Corruption of Scripture, pp. 54-9. 

286 The Text of the New Testament 

it was at Jesus' baptism that he became God's son. It may well be, 
then, that the form of text attested in the majority of witnesses, in this 
case, represents an anti-adoptionistic corruption of the original. 3 "* 

Some passages of the New Testament were modified to stress 
more precisely that Jesus was himself divine. One of the most intrigu- 
ing textual variants in the early part of John's Gospel occurs at the end 
of the prologue, where, according to most witnesses, Christ is said to 
be "the only Son who is in the bosom of the Father" (1.18). Strikingly, 
according to some other witnesses, principally Alexandrian, the 
passage instead calls Christ "the only God who is in the bosom of the 
Father." Whereas many critics continue to regard this as the original 
text, it may well be that the change was implemented by Alexandrian 
scribes who wanted to emphasize Jesus' divinity against those who 
thought that he was merely human. 35 

There were other early Christian groups, including the Marcionites 
and some of the Gnostics, who took just the opposite Christological 
point of view, maintaining that, far from being completely human, 
Christ was fully (and only) divine. In response, some proto-orthodox 
scribes of the second and third centuries occasionally changed their 
texts in order to emphasize that Jesus was in fact a human in every 
way. One important passage occurs in the scene of Jesus' prayer 
before his arrest in Luke's Gospel, where we find the intriguing 
account of the "bloody sweat": "And an angel from heaven appeared 
to him and gave him strength. In his anguish he prayed more 
earnestly, and his sweat became like great drops of blood falling 
down on the ground" (Luke 22.43-4). These verses are absent from 
some of the oldest and best witnesses, including the majority of the 
Alexandrian manuscripts. It is striking to note that the earliest wit- 
nesses attesting the verses are three Church fathers — Justin, Irenaeus, 
and Hippolytus — each of whom uses the verses in order to counter 
Christological views that maintained that Jesus was not a full human 
who experienced the full range of human sufferings. It may well be 
that the verses were added to the text for just this reason, in opposi- 
tion to those who held to a docetic Christology. 36 

34. For further discussion, see Khrman, Orthodox Corruption of Scripture, 
pp. 62-7. 

35. For further discussion, see Khrman, Orthodox Corruption of Scripture, 
pp. 78-82. 

36. For further discussion, see Ehrman, Orthodox Corruption of 
Scripture, pp. 187-94. 

History of the Transmission of the Text 287 

The fullest studies of theological changes to the text have focused 
on such questions of Christology; future studies could profitably ex- 
plore other areas of doctrinal dispute. 

2. Jewish— Christian Relations 

One particularly fruitful area of research over the past 60 years 
has been the study of early Jewish-Christian relations and the rise of 
Christian anti-Judaism. Rooted in the solid investigations of Jules 
Isaac and Marcel Simon and motivated in no small measure by the 
provocative thesis of Rosemary Ruether — that Christianity has by its 
very nature always been anti-Jewish — scholars of both the New 
Testament and later Christianity have produced a voluminous out- 
pouring of literature that discusses the relation of Christianity to its 
Jewish matrix. 37 

How did conflicts with Judaism, evident throughout the first 
three Christian centuries, affect scribes who reproduced the texts of 
Scripture? The question has regrettably not received the extended 
study it deserves. To be sure, even before the Second World War 
scholars had observed that certain manuscripts preserve textual vari- 
ants that are related to the conflicts. Particular mention may be made 
of Heinrich Joseph Vogels and J. Rendel Harris, both of whom 
argued that the anti-Judaic tendencies of Tatian's Diatesseron had 
influenced several of our surviving witnesses. 38 The Curetonian 
Syriac, for instance, modifies the announcement that Jesus will save 
"his people" from their sins (Matt. 1.21) to say that he will save 

37. For bibliography and informed discussion, see John Gager, The 
Origins of Anti-Semitism: Attitudes Toward Judaism in Pagan and Christ- 
ian Antiquity (New York, 1983), pp. 11-34, and more briefly, idem, 
"Judaism as Seen by Outsiders," in Early Judaism and Its Modern Inter- 
preters, ed. by Robert A. Kraft and George W. E. Nickelsburg (Atlanta, 
1986), pp. 99-116. The foundational works include Jules Isaac, Jesus and 
Israel (New York, 1971; French original, 1948); Marcel Simon, Verus Israel: 
A Study of the Relations between Christians and Jews in the Roman Empire 
(135-425) (Oxford, 1986; French orig., 1964); Rosemary Ruether, Faith 
and Fratricide: Ihe 'Theological Roots of Anti-Semitism (New York, 1974). 
Among more recent studies, see Miriam Taylor, Anti-Judaism and Early 
Christian Identity: A Critique of the Scholarly Consensus (Leiden, 1995). 

38. Vogels, Handhuch der Textkritik des Neuen Testaments, 2nd ed. 
(Bonn, 1955; 1st ed. 1923), p. 178; Harris, "Was the Diatesseron Anti- 
Judaic?" Harvard Iheological Review, xviii (1925), pp. 103-9. 

288 The Text of the New Testament 

"the world." So, too, some Syriac and Latin witnesses of the fourth 
Gospel change Jesus' words to the Samaritan woman in John 4,22 to 
indicate that salvation comes "from Judea" rather than "from the 
Jews." Among the most intriguing of the nearly two dozen examples 
that these (and other) scholars have discussed is the omission in 
some manuscripts of Jesus' prayer from the cross, "Father, forgive 
them; for they do not know what they are doing" (Luke 23-34), an 
omission that makes particular sense if Jesus is understood to be 
asking God to forgive the Jews responsible for his crucifixion. 39 

As already mentioned, the most significant study of anti-Jewish 
influences on the text of the New Testament has been Epp's evalu- 
ation of Codex Bezae in Acts. Following earlier suggestions that the 
Western tradition may preserve an anti-Judaic bias, Epp made a 
compelling case that many of the Bezan variants in Acts stand 
against non-Christian Judaism. 40 Even though Epp did not pursue 
the question of Sitz im Leben for this kind of scribal activity, its 
social context in early Christian polemics against Jews is nonethe- 
less clear. Future studies could profitably explore in greater detail 
the significance of this polemical milieu for the textual tradition of 
the New Testament. 41 

3. The Oppression of Women in Early Christianity 

One of the most significant developments in New Testament 
studies over the past 30 years has involved the intensified effort to 
understand the role of women in early Christianity. Those who pur- 
sue the question are by no means unified in their methods or results; 
most notably, some have argued that the Christian tradition is so 
thoroughly and ineluctably patriarchalized that it must be jettisoned 

39. For discussion and bibliography, see Joseph A. Fitzmyer, The 
Gospel According to Luke (X-XX/V) (Garden City, NY, 1985), pp. 1503-4. 

40. For his predecessors, see The Iheological Tendency of Codex 
Bezae, pp. 21-6; in particular, one might mention the study of Menoud, 
op. cit. 

41. On the positive effects of Judaism on the manuscript tradition of 
the New Testament (seen, e.g., in the predisposition among early Chris- 
tians to dispose of texts rather than destroy them), see Colin H. Roberts, 
Manuscript, Society, and Belief in Early Christian Egypt (Schweich Lectures, 
1977; London, 1979). 

History of the Transmission of the Text 289 

altogether, 42 while others have sought to move beyond the biases of 
our sources to reclaim the tradition for themselves. 43 

For the historian concerned with the role of women in earliest 
Christianity, one of the perennial issues relates to the status of 1 Cor. 
14.34-5, a passage that requires women "to be silent in the churches" 
and to "be subordinate." Many scholars have claimed that the passage 
is not Pauline but represents an interpolation, made perhaps by the 
author of (the pseudepigraphic?) 1 Timothy (cf. 2.1— 10). 44 While one 
common objection to the interpolation theory has been the lack of 
manuscript attestation — the passage is present in all of our witnesses — 
Gordon Fee has stressed the text-critical evidence in its support, 
observing that the verses in question occur in a different location in 
Western witnesses (giving the passage the appearance of a marginal 
note incorporated at more or less appropriate junctures). 45 If Fee is 
correct concerning their secondary character, the interpolation may 

42. See, e.g., the provocative discussions of Mary Daly, Beyond God 
the Father: Toward a Philosophy of Women 's Liberation, 2nd ed. (Boston, 
1985); idem, The Church and the Second Sex (New York, 1968). 

43- Most significantly, for the New Testament period, Elizabeth 
Schiissler Fiorenza, In Memory of Her: A Feminist Theological Reconstruction 
of Christian Origins (New York, 1983). An insightful example of feminist 
reconstruction is Antoinette Clark Wire, The Corinthian Women Prophets: A 
Reconstruction Through Paul's Polemic (Minneapolis, 1990). For the second 
century, see the more popular discussion, somewhat less rooted in feminist 
theory, of Elaine Pagels, The Gnostic Gospels (New York, 1979). 

44. Mary Daly objects to those who pursue the status of this passage 
for the sake of exonerating the Apostle Paul: whether he wrote it or not, 
the passage has been used to oppress women and will continue to be 
used in this way {Beyond God the Father, p. 5). At the same time, the ques- 
tion of authorship is important for historians because if Paul did not write 
the verses, then the attitude that they sanction represents a later feature of 
Pauline Christianity. 

45. Gordon D. Fee, The First Fpistle to the Corinthians (Grand Rapids, 
MI, 1987), pp. 699-708. For additional witnesses in support of an omission, 
see Philip B. Payne, "Fuldensis, Sigla for Variants in Vaticanus, and 1 Cor. 
14.34-35," New Testament Studies, xli (1995), pp. 240-62. Payne argues that 
(1) the sixth-century textual scholar Victor, responsible for the revising and 
correcting of the Vulgate manuscript Fuldensis, recognized and indicated 
that 14.34-5 were not original; (2) Clement of Alexandria did not know of 
the verses; and (3 ) the scribe of Codex Vaticanus indicated by means of a 

290 The Text of the New Testament 

show that women came to be oppressed more severely in a later 
period of Pauline Christianity (perhaps around the end of the first 
century) than at the outset. 46 

In an attempt to cast the net somewhat more broadly, Ben 
Witherington has summarized some of the evidence that suggests 
that the scribe of Codex Bezae was intent on de-emphasizing the 
prominent role that women played in the early church, as recorded 
in the narrative of Acts. 47 Labeling such alterations, somewhat inap- 
propriately, "anti-feminist" changes, 48 Witherington observes that in 
Bezae's text of Acts 17.4, Paul's Thessalonian converts are unam- 
biguously "wives of prominent men" rather than "women of promi- 
nence," that the high profile of women is occasionally compromised 
by the insertion of references to their children (Acts 1.14) or to men 
of high profile (Acts 17.12), and that the regular transposition of 
"Aquila" to precede "Priscilla" may intimate the scribe's uneasiness 
with the woman's implicit priority. 

While other scholars have also discussed, in brief order, the 
significance of textual problems for assessing the oppression of 
women in early Christianity, we are still awaiting an extensive and 
rigorous analysis. 

4. Christian Apologia 

A fourth area of the social context of scribes involves the early 
conflicts between Christianity and the Empire, in which Christian 
intellectuals — the apologists — sought to defend the new religion 
against the attacks of its cultured and/or politically motivated 

siglum resembling an umlaut in the margin of the text that the verses 
were omitted in part of the manuscript tradition known to him (see also 
J. Edward Miller's conclusion found at the end of n. 10 on p. 259). 

46. More severely because they were already treated differently from 
men in the early period; they were required, for example, to wear veils 
when praying or prophesying (1 Cor. 11.2-10). Interestingly, Fee's 
arguments have not been accepted by Antoinette Wire, a recent feminist 
historian who attempts to reconstruct the situation in Corinth {The 
Corinthian Women Prophets, pp. 229-32). On other developments in the 
Pauline communities, see Jouette M. Bassler, "The Widows' Tale: A Fresh 
Look at 1 Tim 5:3-16," Journal of Biblical Literature, ciii (1984), pp. 23-41. 

47. B. Witherington, "The Anti-Feminist Tendencies of the 'Western' 
Text in Acts," Journal of Biblical Literature, ciii (1984), pp. 82-4. 

48. The label is anachronistic and misleading, since these changes are 
not directed against feminists (a modern intellectual category). 

History of the Transmission of the Text 291 

despisers. These conflicts, too, affected the text of the New Testa- 
ment, as has been recently shown by the study of Wayne Kannaday. 49 

By A.D. 180 or so, some of the better-informed pagan opponents 
of Christianity, such as the Middle-Platonist Celsus, had read the 
Gospels and used their portrayals of Jesus as weapons against the 
Christians. A heated debate commenced — in literary circles, at least — 
over whether the things Jesus said and did were in fact appropriate 
to one who was revered as the Son of God. The background to 
these debates lay in the widespread notion throughout the Mediter- 
ranean that divine men occasionally roamed the earth. There were, 
of course, numerous stories about other superhuman individuals, 
who, like Jesus, were also said to have been supernaturally born; 
to have performed miracles such as healing the sick, casting out 
demons, and raising the dead; and to have been exalted to heaven to 
live with the gods. These other individuals were also sometimes 
called "sons" of God. 30 

Based on the fragmentary evidence at our disposal, it appears 
that there were general expectations of what such a person would be 
like within the broader culture of the Greco-Roman world. Part of 
the confrontation between pagans and Christians, at least in the 
rarified atmosphere of the apologetic literature, involved determin- 
ing whether Jesus carried himself with the dignity and deportment 
expected of a son of God. Pagan critics like Celsus argued on the 
contrary that Jesus was a fraud who did not benefit the human race 
and that, as a consequence, he was not a true son of God but a 
deceiver, a worker of dark craft, a magician. 51 

These debates over Jesus' identity and the appropriateness of his 
being designated as the Son of God occasionally affected the trans- 
mission of texts of the New Testament. One of the most interesting 
cases involves a passage in which some of the earliest papyri attest 
a corruption. This is the account in Mark 6.3, in which, according to 
most witnesses, the townspeople of Nazareth identify Jesus as the 
"carpenter, the son of Mary." We know that Celsus himself found this 
identification significant, possibly (though not certainly) because it 
situated Jesus among the lower classes and thereby showed him not 

49. W. Kannaday, Apologetic Discourse and the Scribal Tradition 
(Atlanta, 2004). 

50. For a study of this question from the perspective of early Christian 
apologia, see Eugene Gallagher, Divine Man or Magician: Celsus and 
Origen on Jesus (Chico, CA, 1982). 

51. See Gallagher, Divine Man or Magician. 

292 The Text of the New Testament 

to be worthy of divine stature. Origen's response may have been 
disingenuous, although there is no way to know for certain; he 
claims that none of the Gospels of the church provides this 
identification. Possibly all of Origen's manuscripts of Mark agreed 
with p 45 , f 13 , and 33 in changing 6.3 to identify Jesus as "the son of 
the carpenter," rather than "the carpenter"; or possibly he had for- 
gotten the passage in Mark. In any event, given the second-century 
modification of the text — that is, its change precisely in the period 
when Jesus' own socioeconomic status had become an issue for 
apologists — we might be inclined to think that it was precisely the 
apologetic impulse that led to the corruption. 52 

Another example involves the modification of Mark 1.41. 53 In 
every Greek manuscript except Codex Bezae, Jesus responds sym- 
pathetically to a leper's request for healing: "and moved with pity, 
he stretched out his hand and touched him." More surprising is the 
response recorded in Codex D (along with several Latin allies), 
where Jesus' compassion is turned into wrath: "and moved with 
anger he stretched out his hand and touched him." 

The reading is obviously not well attested, and for this reason it 
has been rejected over the years by the majority of critics and com- 
mentators. 54 But here, one must press hard the transcriptional issue: if 
the oldest form of the text had indicated that Jesus reacted to this poor 
soul with compassion, why would any scribe modify it to say that he 
became angry? But if the text had originally mentioned Jesus' wrath, 
it is quite easy to imagine scribes taking offense and modifying the 
text accordingly. Indeed, the scribal offense may not have involved so 
much a general puzzlement as a specific fear, namely, that the pagan 
opponents of Christianity, like Celsus, who were known to be perus- 
ing the Gospels for incriminating evidence against the divine founder 
of the faith, might find here ammunition for their charges. 

52. This has to be inferred from Origen's reference to Celsus' work, in 
light of his overall polemic (on which, see Gallagher, Divine Man or Magi- 
cian); see Contra Celsum, vi, 36. 

53- For a fuller discussion, see Bart D. Ehrman, "A Leper in the Hands 
of an Angry Jesus," in New Testament Greek and Exegesis: Essays in Honor 
of Gerald Hawthorne, ed. by A.M. Donaldson and T.B. Sailors (Grand 
Rapids, MI, 2003), pp. 77-98. 

54. For an insightful discussion of the problem and a persuasive argu- 
ment for the more difficult reading, see Joel Marcus, Mark: A New Transla- 
tion with Interpretation and Commentary (New York, 2000). 

History of the Transmission of the Text 293 

There is other evidence to support the view that the text origi- 
nally indicated that Jesus became angry when approached by this 
leper. We do well to recall that the scribes of our surviving manu- 
scripts were not the first copyists of Mark's Gospel. Strictly speaking, 
the earliest surviving copies were made by Matthew and Luke (even 
though these copies were modified far beyond what later scribes 
would dare to do). If one accepts any form of Marcan priority, as 
most scholars continue to do (despite some renewed attacks in 
recent years), it may be profitable in an instance such as this to see 
what the other Synoptics do with the passage. 

It is striking that both Matthew and Luke retain the story for their 
own accounts, by and large reproducing it verbatim. But they both 
also omit the participle in question (cmXay%vio0£ig or oygiaOeig). It 
must be conceded that Matthew and Luke each modifies this story of 
Mark to his own ends in other ways as well — as they do with most 
of Mark's other stories. Rarely, however, do they change their Marcan 
source in the same ways. The so-called minor agreements between 
Matthew and Luke have traditionally caused the largest problems for 
the proponents of Marcan priority. What is remarkable is that the 
majority of these minor agreements appear to be agreements in 
omission, and most of them can be explained by positing something 
offensive or puzzling in Mark's account — that is, something that may 
have appeared offensive or puzzling independently to more than 
one redactor. 55 

It must be pointed out, in this connection, that in no other in- 
stance in which Jesus is said to feel compassion (.OTikayxvvodeig) in 
Mark's Gospel do both Matthew and Luke change it. 56 Why would 
the participle be omitted here? It would make sense if in fact Mark's 
text, as both Matthew and Luke had it, did not indicate that Jesus felt 
compassion (ojtXayxviodeig) but wrath {dygiodeig). But does it 
make sense in Mark's own narrative for Jesus to become angry prior 
to healing the man? 57 Jesus does become angry at other times in 
Mark's Gospel (3.5, 10.41; both of these are edited out of the stories 
by Matthew and Luke!). With respect to the present context, as 
several commentators have suggested, it is difficult to explain the 

55. As one example among many, see the omission of "You shall not 
defraud" from the list of the commandments in Mark 10.19. 

56. Mark 6.34, 8.2. 

57. See the discussion in Marcus, op. cit. 

294 The Text of the New Testament 

severity of 5.43 apart from some such intimation earlier: "and 
censuring him, he immediately cast him out" (xai ifi^Qt/urjod/uevog 
avrqj evOvq e^efiakev avtov).^ Finally, the change of the text from 
an angry to a compassionate Jesus makes sense in light of the apolo- 
getic efforts reflected in later authors such as Origen, one of whose 
burdens was to show not only that Jesus caused no real offense but 
that his presence among humans wrought great benefits. An empha- 
sis on Jesus' compassion would significantly aid in the rebuttal of the 
charges against Jesus made by opponents like Celsus. 

Other textual variants appear to serve similar apologetic ends. 
This certainly appears to be the impulse behind the changed word 
order of Luke 23.32, where the original statement that Jesus was 
crucified with "two other criminals" is modified in the majority of 
witnesses to read "two others, who were criminals." An even more 
effective change was made in some of the versional witnesses (c, e, 
sy s ), where the offensive term is omitted altogether so that Jesus is 
crucified along with "two criminals." 

Apologetic impulses may also lie behind several changes in the 
texts of the Gospels that appear to prevent Jesus, the all-knowing 
Son of God, from making factual mistakes in what he says. This may 
be the case, for example, in the deletion of the problematic state- 
ment in Mark 2.26 that David's entry into the temple to partake of the 
showbread took place "when Abiathar was high priest" (when in fact 
it was his father Ahimelech who was the high priest at the time; see 1 
Sam. 21.1-7) and the statement of Jesus to the high priest in Mark 
14.62 that he, the high priest, would see the son of man "coming 
with the clouds of heaven" (since, in fact, he died long before this 
event to end all events). Both passages were changed by scribes, 
possibly in an attempt to prevent the reader from concluding that 
Jesus had made a mistake. 

5. Christian Asceticism 

A fifth area of interest involves the impact of ascetic Christianity 
on the copyists of Scripture, an impact that was comparatively slight 
but nonetheless noteworthy. One of the most famous instances of 

58. It would make much less sense for scribes to change 5.41 in light 
of 5.43, which they would have copied afterward. 

History of the Transmission of the Text 295 

an ascetically oriented alteration 59 comes in Mark's account of Jesus' 
words to his disciples after they proved unable to cast out a partic- 
ularly difficult demon: "This kind can come out only through 
prayer" (Mark 9.29)- A large number of witnesses make an important 
addition to Jesus' words: this kind of demon "can come out only 
through prayer and fasting." Now the ascetic life is shown to be 
necessary for one to overcome the Satanic forces of evil unleashed 
on this world. 

Two other examples come in readings whose presence in our 
Syriac sources was earlier in the twentieth century attributed to the 
encratitic tendencies of Tatian's Diatessaron. The first appears in 
Luke 2.36, where in the Sinaitic Syriac we learn that the prophet 
Anna enjoyed marital bliss not for "seven years" but only for "seven 
days"; the other comes in Matt. 22.4, where the same manuscript 
leaves the oxen and fatted calves off the menu of the divine marriage 
feast. One is reminded of the diet of John the Baptist, described by 
the author of the Gospel of the Ebionites; by changing one letter and 
adding another, this proponent of vegetarian cuisine served the 
Baptist pancakes (eyxQideg) rather than locusts (dxgideg). 60 Scribes 
who changed their texts in this way appear to have been intent on 
the renunciatory Christian life. 

6. The Use of Magic and Fortune-Telling 

in Early Christianity 

One of the more fascinating areas of biblical scholarship involves 
the role of magic in the early Church. Not everyone who works in 
this area agrees even on the most basic of questions, such as the de- 
finitions of magic and religion and how, or whether, they can be 
neatly differentiated from one another. 61 Nonetheless, a number of 

59. For an overview of the rise of early Christian asceticism, see 
Elizabeth Clark, Reading Renunciation: Asceticism and Scripture in Early 
Christianity (Princeton, 1999), especially Chapter 1 and the literature she 
cites there. 

60. See Epiphanius, Panarion, xxx, 13, 4-5. 

61. For a useful discussion, see David E. Aune, "Magic in Early 
Christianity," Aufstieg und Niedergang der Romischen Welt, 2.23.2, 
pp. 1506-16. 

296 The Text of the New Testament 

creative and insightful studies have been produced in recent years, 
some dealing with the role of magic in the life of Jesus, others with 
its portrayal in the New Testament narratives, yet others with its pop- 
ularity among early Christians. 62 

Even though there do not appear to be many (if any) textual vari- 
ants that reflect the practice of magic in early Christianity, we know 
from literary sources of the fourth century and later that New Testa- 
ment manuscripts themselves were sometimes used for apotropaic 
magic — for example, worn around the neck or placed under a pillow 
to ward off evil spirits. 63 Among the papyri discovered and analyzed 
during the past 50 years are several that were beyond any doubt 
made and used as amulets: they are small in size, often a single sheet 
folded over, sometimes provided with or tied together with a string, 
and normally inscribed with texts that could prove useful for warding 
off evil spirits or for effecting healings; — the Lord's Prayer, for exam- 
ple, or a healing narrative. 64 A full discussion of these scriptural 
amulets awaits further study. 65 

62. The best overview, with extensive bibliography, is Aune, op. cit. 
For the role of magic in the life of Jesus, see especially the provocative 
studies of Morton Smith, Jesus the Magician (San Francisco, 1978) and, 
more extensively, Clement of Alexandria and a Secret Gospel of Mark 
(Cambridge, MA, 1973). For an interesting assessment of the portrayal of 
magic in the New Testament, see Susan R. Garrett, The Demise of the Devil: 
Magic and the Demonic in Luke's Writings (Minneapolis, 1989). Most re- 
cent studies have been inspired by the publication of magical texts from 
the Greco-Roman world. For English translations, see Hans Dieter Betz, 
ed., The Greek Magical Papyri in Translation (Chicago, 1986). 

63. E.g., John Chrysostom, Horn. 19.4; see the discussion in 

R. Kaczynski, Das Wort Gottes in Liturgie und Alltag derGemeinden des 
Johannes Chrysostomus (Freiburg-im-Breisgau, 1974). 

64. For arguments and examples, see E. A. Judge and Wilbur Picker- 
ing, "The Magical Use of Scripture in the Papyri," in Perspectives on Lan- 
guage and Text, ed. by Edgar W. Conrad and Edward G. Newing (Winona 
Lake, IN, 1987); Roberts, Manuscript, Society, and Belief , pp. 82-3. In his 
original edition of the Greek magical papyri, K. Preisendanz classified 38 
of the 107 available texts as Christian (Papyri Graecae Magicae, 2nd ed. by 
A. Henrichs [Stuttgart, 19731); according to Judge (op. cit., p. 341), 15 of 
these 38 "make conscious use of scriptural material." 

65. See especially Judge and Pickering, op. cit., and the bibliography 
cited there. 

History of the Transmission of the Text 297 

Closely connected with the question of magic is the practice of 
fortune-telling in the ancient world, on which a number of interesting 
studies have been produced, particularly with respect to the Sortes 
Astrampsychi and others of the so-called Books of Fate. 66 Little, how- 
ever, has been written about the use of fortune-telling in early Chris- 
tianity, perhaps due to a dearth of evidence. 67 Indeed, some of the 
most intriguing evidence happens to derive from the manuscript 
tradition of the New Testament. 

Among noteworthy features of New Testament manuscripts 
(besides those mentioned on pp. 92-4) is the presence in several 
early Greek copies of the Gospel according to John of what are 
called EQjunvela. The scribes of eight papyrus or parchment copies of 
John, dating from the third (or fourth) century to the eighth century 
(namely, p 55 , p 59 , p 63 , p 76 , p 80 , 0145, 68 0210, and 0256), 69 have followed 
a certain stereotyped pattern. Each page is arranged so that the 
Scripture text, which does not fill the page, is followed by the word 
igfir/vela, centered as a title, and this is followed by a sentence or a 
phrase. What are these "hermeneiai"? 

On the basis of the title, the opinio communis has been that the 
sentences are a kind of rudimentary commentary on Scripture. That 
this opinion, however, is false becomes apparent as soon as one 

66. For a brief description, see T. C. Skeat, "An Early Mediaeval 'Book 
of Fate': The Sortes XII Patriarcharurn. With a note on 'Books of Fate' in 
General," Mediaeval and Renaissance Studies, iii (1954), pp. 41-54. On the 
Sortes Astrampsychi, see the overview of G, M. Browne, "The Composition 
of the Sortes Astrampsychi, 11 Bulletin of the Institute of Classical Studies, 
xvii (1970), pp. 95-100; for basic bibliography, see idem, "The Sortes 
Astrampsychi and the Egyptian Oracle," Texte und Textkritik: Eine 
Aufsatzsammlung, ed. by Jiirgen Dummer (Berlin, 1987), p. 71. 

67. On broader issues related to prophecy in early Christianity in 
conjunction with divination and oracles in the Greco-Roman world, see 
especially David E. Aune, Prophecy in Early Christianity and the Ancient 
Mediterranean World (Grand Rapids, MI, 1983). 

68. Unfortunately, this parchment leaf cannot now be found, but H. 
von Soden described it as having the hermeneia in red ink {Die Schriften 
des Neuen Testaments, i [Berlin, 1902], p. xi). 

69. Besides these eight texts, a ninth (p 60 ) probably also had 
eg/invEia, though owing to its fragmentary condition, none has been pre- 
served. For 12 fragments of a sixth-century Gospel according to John in 
Coptic (Sahidic), supplied with egftnveta in Greek, see W. E. Crum in 
Proceedings of the Society of Biblical Archaeology, xxvi (1904), pp. 174-6. 

298 The Text of the New Testament 

notices that the "comments" are totally irrelevant to the passage with 
which they share the page. What then can they be? In seeking an 
answer to this question, one observes that the so-called comments 
are similar in form and, in some cases, in substance to the series of 
short apophthegms standing one per page in the lower margin of the 
pages which contain the text of Mark 1.1-10.22 (folios 285 b to 321a) 
in the fifth-century manuscript Codex Bezae. Written in a wretched, 
scrawling Greek hand of perhaps the ninth or tenth century, each of 
these 69 short statements in Bezae is preceded by the word EQfirjvEia 
or its abbreviation. 70 Typical of phrases preserved in the eight frag- 
mentary copies of John are the following on folios 291 b to 293£> of 
Bezae: "Expect a great miracle," "You will receive joy from God," 
"From pain to joy," "After ten days it will happen," and "What you 
seek will be found." 

The existence of ancient and medieval 71 fortune-telling manu- 
als, some of which show parallels with the items mentioned above 
(including in some cases the prefatory use of the word EQfirjveld), 
have led to the proposal that the equipment in the margins of Codex 
Bezae and the manuscripts of John must have been used for the 
purpose of divination (Sortes sanctorum), that is, the telling of 
fortunes. 72 A number would be selected, perhaps by throwing dice, 
and then the pages of the Gospel codex would be turned until the 
sentence that corresponded to the number was found. 

70. For a list of these 69 items, see Frederick H. [A.] Scrivener, Bezae 
Codex Cantabrigiensis (Cambridge, 1864; repr. Pittsburgh, 1978), pp. 451-2; 
for reproductions of folios of Mark in Codex Bezae, see H. J. Vogels, 
Codicum Novi Testamenti specimina (Bonn, 1929), reproduced in 

H. Zimmermann, Neutestamentliche Methodenlehre (Stuttgart, 1968). 

71. A Byzantine book of fate ('Pixxokdyiov), preserved in two Greek 
manuscripts, presents a list of 38 short passages from the four Gospels (be- 
ginning with John 1.1), each followed by a fortune prefaced by the word 
eQftrjVEia (see F. Drexl, "Ein griechischen Losbuch," Byzantinische 
Zeitschrift, xli [1941], pp. 311-8). 

72. See B. M. Metzger, "Greek Manuscripts of John's Gospel with 
TIermeneiai,' " in Text and Testimony, Essays on New Testament and 
Apocryphal Literature in Honour of A. F.J. Klijn, ed. by Tj. Baarda et al. 
(Kampen, 1988), pp. 162-9. Moreover, information from Armenian and 
Georgian manuscripts has been provided by B. Outtier, "Les Prosermeneiai 
du Codex Bezae," in Codex Bezae; Studies from the Lunel Colloquium June 
1994, ed. by D. C. Parker and C.-B. Amphoux (Leiden, 1996), pp. 74-8. 

History of the Transmission of the Text 299 

Thus, it is evident that some Christians ascribed special powers 
to the manuscripts of Scripture themselves: they could be used not 
only for purposes of apotropaic magic (the amulets) but also to in- 
fluence, or at least predict, one's future. This then is a unique kind 
of evidence for the historian of the period: it can tell us about the 
role of sacred texts for the life of ordinary Christians — as opposed, 
that is, for the lives of the Christian elite who produced our literary 
evidence. Here again, however, a full study of the phenomenon 
remains a desideratum. 


The Practice of New Testament 
Textual Criticism 

I. Basic Criteria for the Evaluation 
of Variant Readings 

Perhaps the most basic criterion for the evaluation of variant read- 
ings is the simple maxim "choose the reading that best explains the 
origin of the others." We all follow this common-sense criterion 
when confronted with errors and "variant readings" in modern 
printed books. For example, two editions of John Bunyan's classic- 
ize Pilgrim 's Progress diverge in the story of Christian's finding and 
using a key by which he is able to make his escape from Doubting 
Castle. One edition reads "The lock went desperately hard," while 
the other reads "The lock went damnable hard." Which is the original 
reading and which has been altered? Did Bunyan write "desperately" 
and a modern editor change it to "damnable" for some inexplicable 
reason? Or did Bunyan write "damnable" (using the word in its non- 
profane sense) and someone subsequently altered it in order to 
remove what was deemed to be an offensive expression? There can 
surely be no doubt what the answer is. 1 

1. The example given above has been simplified; at least three differ- 
ent modifications of the original reading have been introduced by editors 
or printers. Besides "desperately hard," other copies read "extremely hard" 


The Practice of New Testament Textual Criticism 301 

Another criterion that we instinctively recognize to be basic is 
that the reconstruction of the history of a variant reading is prereq- 
uisite to forming a judgment about it. For example, in the earlier 
printings of the second edition of the unabridged Webster's New 
International Dictionary of the English Language (Springfield, MA, 
1934) there stands the following entry: 

dord (dord), n. Physics & Chem. Density. 

Now, it is a fact that there is no English word dord; its presence in 
this venerable dictionary is the result of what may be called an acci- 
dental "scribal error." As was acknowledged later by the publishers, 
the entry originated in the confusion of the abbreviation, given 
as both a lower-case letter and an upper-case letter, of the word 
density and was intended to stand thus: 

d. or D., Physics & Chem. Density. 

Not noticing the periods, someone took the collocation of letters as 
a word and called it a noun. The remarkable thing is that the error 
escaped detection for more than a decade, during which the volume 
was reprinted several times. 

Another example of a clerical mistake, this one occurring in the 
highly esteemed Who's Who in America, arose because of incompe- 
tent judgment. The first time that the biography of Thomas Mann 
appeared in this distinguished cyclopedia of famous persons, he was 
given, quite gratuitously, a middle name. In the volume for 1939, the 
entry reads, in bold-face type, "Mann, Thomas Schriftst"; in sub- 
sequent volumes, however, "Schriftst" is lacking. Which form of 
the name is correct? An examination of the volume Werlst's, which 
is the German counterpart of Who's Who, discloses that "Schriftst" is 
the customary abbreviation of the German word for "author" 
(Schriftsteller). Obviously, someone who prepared the biographical 
sketch for the American volume mistakenly took the abbreviation of 
Mann's occupation to be his middle name. 

The two criteria mentioned earlier are capable of very wide 
application and include by implication a great many other sub- 
sidiary criteria. It will be useful, however, to specify in more precise 
detail the various considerations that scholars take into account in 

or "very hard"; for the authentic text, reference may be made to James B. 
Wharey's collation of the first 1 1 editions of The Pilgrim 's Progress (Oxford, 
1928), all of which read "damnable hard." 

302 Thh Tkxt of the Nkw Testament 

evaluating variant readings of New Testament witnesses. It is usual 
to classify these criteria in terms of (1) external evidence and (2) in- 
ternal evidence; the latter involves what Hort termed "transcrip- 
tional probabilities" and "intrinsic probabilities." (Here, the student 
should re-read the account of the principles underlying Westcott 
and Hort's edition, pp. 174-81 as well as the summary of IS. H. 
Streeter's subsequent contributions to textual theory, pp. 214-8.) 
The following is a list of the chief considerations and criteria that the 
textual critic takes into account when evaluating variant readings in 
the New Testament. 

1. External evidence, involving considerations bearing upon: 

a. The date of the witness. (Of even greater importance than the 
age of the document itself is the date of the type of text that it 
embodies. The evidence of some minuscule manuscripts [e.g., 
33, 81, and 1739] is of greater value than that of some of the 
later or secondary majuscules.) 

b. The geographical distribution of the witnesses that agree in sup- 
porting a variant. (One must be certain, however, that geographi- 
cally remote witnesses are really independent of one another. 
Agreements, for example, between Old Latin and Old Syriac 
witnesses may be due to influence from Tatian's Diatessaron.) 

c. The genealogical relationship of texts and families of witnesses. 
(Witnesses are to be weighed rather than counted. Further- 
more, since the relative weight of the several kinds of evidence 
differs for different kinds of variant, there can be no merely 
mechanical evaluation of the evidence.) 

2. Internal evidence, involving two kinds of probability: 

a. Transcriptional probabilities depend on considerations ofpale- 
ograpbical details and the habits of scribes? Thus: 

i. In general, the more difficult reading is to be preferred, 
particularly when the sense, on the surface, appears to 

2. For discussions of scribal tendencies, see James R. Roysc, "Scribal 
Habits in the Transmission of New Testament Texts," in The Critical Study 
of Sacred Texts, ed. by W. D. O'Flanerty {Berkeley Religious Studies Series; 
Berkeley, 1979), pp. 139-61; idem, "Scribal Tendencies in the Transmission 
of the Text of the New Testament," in The Text of the New Testament in 
Contemporary Research: Essays on the Status Quaestionis, ed. by Bart D. 

The Practice of New Testament Textual Criticism 303 

be erroneous but, on more mature consideration, proves 
to be correct. (Here, "more difficult" means "more difficult 
to the scribe," who would be tempted to make an emen- 
dation. The characteristic of most scribal emendations is 
their superficiality, often combining "the appearance of 
improvement with the absence of its reality." 3 Obviously, 
the category "more difficult reading" is relative, and a point 
is sometimes reached when a reading must be judged to 
be so difficult that it can have arisen only by accident in 

ii. In general, the shorter reading is to be preferred, except 
where parablepsis arising from homoeoteleuton may have 
occurred or where the scribe may have omitted material 
that he deemed to be superfluous, harsh, or contrary to 
pious belief, liturgical usage, or ascetical practice. (Compare 
Griesbach's fuller statement of this criterion, pp. 166-67 

ill. Since scribes would frequently bring divergent passages into 
harmony with one another, in parallel passages (whether 
involving quotations from the Old Testament or different 
accounts of the same event or narrative) that reading is to be 
preferred which stands in verbal dissidence with the other. 

iv. Scribes would sometimes replace an unfamiliar word with 
a more familiar synonym, alter a less refined grammatical 
form or less elegant lexical expression in accordance with 
Atticizing preferences, or add pronouns, conjunctions, and 
expletives to make a smooth text. 

b. Intrinsic probabilities depend on considerations of what the 
author was more likely to have written, taking into account: 

i. the style, vocabulary, and theology of the author through- 
out the book, 
ii. the immediate context, 
ill. harmony with the usage of the author elsewhere, 

lihrman and Michael W. Holmes (Grand Rapids, MI, 1995), pp. 239-52; 
Peter N. Head, "Observations on Karly Papyri of the Synoptic Gospels, 
especially on the 'Scribal Habits,'" Biblica, lxxi (1990), pp. 240-7. 

3. B. F. Westcott and F. J. A. Hon, The New Testament in the Original 
Greek, ii (London, 1881; 2nd ed., 1896), p. 27. 

304 The Text of the New Testament 

iv. the Aramaic background of the teaching of Jesus, 
v. the priority of the Gospel according to Mark, and 
vi. the influence of the Christian community upon the formu- 
lation and transmission of the passage in question. 

Not all of these criteria are applicable in every case. 4 The critic 
must know when it is appropriate to give primary consideration to 
one type of evidence and not to another. Since textual criticism is 
an an as well as a science, it is understandable that in some cases 
different scholars will come to different evaluations of the signifi- 
cance of the evidence. This divergence is almost inevitable when, as 
sometimes happens, the evidence is so divided that, for example, the 
more difficult reading is found only in the later witnesses. 

One of the perennial dangers that confront scholars in every dis- 
cipline is the tendency to become one-sided and to oversimplify 
the analysis and resolution of quite disparate questions. In textual 
criticism, this tendency can be observed when a scholar, becoming 
enamored of a single method or criterion of textual analysis, applies 
it more or less indiscriminately to a wide variety of problems. For 
example, at the beginning of the twentieth century Adalbert Merx 
devoted three learned volumes to the attempt to prove that the 
Western text is closest to the original and that its best representative 
is the Sinaitic Syriac palimpsest. 5 So, too, Adolf von Harnack, believ- 
ing that the principles of New Testament criticism needed to be 
revised, suggested that the Latin Vulgate had been largely over- 
looked in the arsenal of the critic's tools. 6 Though it may well be that 
some scholars had not given due consideration to Jerome's contri- 
butions to textual criticism, von Hamack's proposal to accord the 

4. For an overview of some of the problems, see Jacobus Petzer, 
"Author's Style and the Textual Criticism of the New Testament," Neotesti- 
mentica, xxiv (1990), pp. 185-97. 

5. Adalbert Merx, Die vier kanonischen Evangelien nach ihrem dltesten 
bekannten Texte; zweiter Theil, Erlduterungen (Berlin, 1902, 1905, 1911). 

6. Adolf von Harnack, Zur Revision der Prinzipien der neutesta- 
mentlichen Textkritik die Bedeutung der Vulgatafiir den Text der katholis- 
chen Briefe und der Anteil des Hieronymus an dem Ubersetzungswerk 
(Beitrdge zur Einleitung in das Neue Testament, 7. Teil; Leipzig, 1916); 
idem, "Studien zur Vulgata des Hebraerbriefs," Sitzungsberichte der 
Preussischen Akademic, 1920, pp. 179-201 (Studien zur Gescbichte des 
Neuen Testaments und der Alten Kirche, i, ed. by Hans Lietzmann [Berlin, 
19311, pp. 191-234). 

The Practice of New Testament Textual Criticism 305 

Vulgate text a preponderant weight in the evaluation of variant read- 
ings found many adverse critics, including several Roman Catholic 
scholars. 7 Similarly, von Soden's extravagant estimate of the influ- 
ence that Marcion and Tatian exerted upon the text of the New 
Testament and A. C. Clark's repeated appeal to the longer text 
are generally regarded today as warnings against a one-sided and 
unwarranted oversimplification of the evidence. 8 

II. The Process of Evaluating 
Variant Readings 

To teach another how to become a textual critic is like teaching 
another how to become a poet. The fundamental principles and 
criteria can be set forth and certain processes described, but the 
appropriate application of these in individual cases rests upon the stu- 
dent's own sagacity and insight. With this caveat in mind, the beginner 
will know how to estimate the following simplified description of text- 
critical methodology. 

As a preliminary step in analyzing and evaluating the evidence 
found in a critical apparatus, the several variant readings should be 
set down in a list, each with its supporting witnesses. This will help 
one to see clearly the point at issue and whether the documents have 
two principal readings or more. 

1. External Evidence 

In evaluating the evidence, the student should begin with exter- 
nal considerations, asking which reading is supported by the earliest 
manuscripts and by the earliest type of text. Readings that are early 

7. See, e.g., Michael Hetzenauer, De recognitione principiorum criticae 
textus Novi Testamenti secundum Adolfum de Harnack (Ratisbonae, 1921 ; 
repr. from Lateranum, 1920, no. 2); J. Belser, "Zur Textkritik der Schriften 
des Johannes," Theologische Quartalschrift, xcviii (1916), pp. 145-84. 

8. Housman derided at this type of textual critic: 

We must have no favourite method. An emendator with one method is as 
foolish a sight as a doctor with one drug. The scribes knew and cared no more 
about us and our tastes than diseases care about the tastes of doctors; they 
made mistakes not of one sort but of all sorts, and the remedies must be of all 
sorts too. (A. E. Housman, ed., M. Manila Astronomicon, liberprimus 
[Cambridge, 19031, pp. If ii f.) 

306 The Text of the New Testament 

and supported by witnesses from a wide geographical area have a 
certain initial presumption in their favor. On the other hand, read- 
ings that are supported by only Koine or Byzantine witnesses (Hort's 
Syrian group) may be set aside as almost certainly secondary. 9 The 
reason that one is justified in discarding the Koine is that it is a later 
text type, formed on the basis of earlier types. Despite the fact that 
it appears in a large majority of Greek manuscripts (for it was 
adopted, with subsequent modifications, as the received text of the 
Greek Orthodox Church), the abundance of witnesses numerically 
counts for nothing in view of the secondary origin of the text type as 
a whole. 10 

To facilitate the process of ascertaining which types of text 
support the several variant readings, the student should become 
thoroughly familiar with the following tables of witnesses. One must 
beware, however, of supposing that these text types are static and 
exactly defined entities; on the contrary, each type involves a 
process 11 of textual development which, though distinctive and char- 
acteristic as a whole, cannot be isolated within precisely determined 

Koine or Byzantine Witnesses 

Gospels: A, E, F, G, H, K, P, S, V, W (in Matthew and Luke 
8.13-24.53), n, V (in Luke and John), Q, and most minuscules. 
Acts: H a , L ap , P a , 049, and most minuscules. 

9. Theoretically, it is possible that the Koine text may preserve an 
early reading that was lost from the other types of text, but such instances 
are extremely rare (one example is discussed on pp. 339-40). See Harry 
Sturz, The Byzantine Text Type and New Testament Textual Criticism, 2nd 
ed. (Nashville, 1984). For a survey of previous evaluations of such read- 
ings, see the chapter on the Lucianic recension in B. M. Metzger, Chapters 
in the History of New Testament Textual Criticism (Grand Rapids, MI, 1963), 
pp. 1—41. 

10. See the discussion of the "majority text debate," pp. 218-22. On 
the rise of the Byzantine textual tradition, see Klaus Wachtel, Der byzanti- 
nische Text der katholischen Briefe: eine Untersuchung zur Entstehung der 
Koine des Neuen Testaments (Berlin, 1995). 

11. See E. C. Colwell, ".Method in Establishing the Nature of Text-Types 
of New Testament Manuscripts," in Studies in Methodology in Textual Criti- 
cism of the New Testament (Grand Rapids, MI, 1969), pp. 44-55, especially 
pp. 53 f. 

The Practice of New Testament Textual Criticism 307 

Epistles: L ap , 049, and most minuscules. 
Revelation: 046, 051, 052, and many minuscules 


Pre-Koine Types of Text 

The forms of text that antedate the Koine or Byzantine text 
include the Western group, the so-called Caesarean, and the Alexan- 
drian (Hort's "Neutral"). 13 


A type of text of the Greek New Testament marked by a distinc- 
tive cluster of variant readings was named the "Western" text 
because the chief witnesses to it were thought to be of Western 
provenance, that is, some Greco— Latin manuscripts (e.g., Codex 
Bezae), the Old Latin, and quotations in the Latin fathers. It is now 
acknowledged that this type of text is not confined to the West; some 
of its variant readings appear also in Eastern versions, such as the 
Sinaitic Old Syriac and the Coptic. Consequently, when the designa- 
tion continues to be used by textual critics, it is more as a proper 
name than as a geographical term. 

Although some have held that the Western text was the deliber- 
ate creation of an individual or several individuals who revised an 
earlier text, 14 most scholars do not find this type of text homogeneous 
enough to be called a textual recension; it is usually considered to 
be the result of an undisciplined and "wild" growth of manuscript 

12. The Byzantine text of the Book of Revelation is less homogeneous 
than it is in other books of the New Testament, for the Greek Orthodox 
Church has never included readings from the Apocalypse in its lectionary 
system — a system that exerted a stabilizing influence on the Byzantine text 
of other books of the New Testament. 

13. For fuller descriptions of each of these three pre-Koine types of 
text as well as lists of witnesses that support each type in the several nat- 
ural divisions of the New Testament (Gospels, Acts, Pauline Epistles, 
Catholic Epistles, and Revelation), see M.-J. Lagrange, Critique textuelle, ii, 
la critique rationnelle (Paris, 1935). Somewhat different lists are given in 
the preface of August Merk's Novum Testamentum graece et latine, 8th ed. 
(Rome, 1957). 

14. So, e.g., James Hardy Ropes, The Text of Acts {The Beginnings of 
Christianity, part I, vol. iii; London, 1926), pp. viii ff.; W. H. P. Hatch, 
'the "Western" Text of the Gospels (F.vanston, IL, 1937). Ropes suggests 
that the preparation of the Western text was involved in the work of 
forming the primitive canon of the New Testament. 

308 The Text of the New Testament 

tradition and translational activity. 15 A marked characteristic of this 
text is the love of paraphrase, resulting in clearly secondary features 
of addition, omission, substitution, and "improvement" of one kind or 

Because the Western type of text was used by such second- and 
early third-century authors as Marcion, Justin (and probably Tatian), 
Heracleon, Irenaeus, and Tertullian, most scholars date the emer- 
gence of the Western text to the mid-second century or shortly there- 
after. 16 But they also, as Martini has put it, "leave the door open to 
an appreciation of the presence of particular readings in which D or 
other 'Western' witnesses have, perhaps, preserved the most ancient 
reading." 17 

Various theories of the origin of the Western type of text have 
been proposed. Westcott and Hort considered it to have arisen as a 
deliberate second-century revision. Others tried to explain it as the 

15. The study of B. Aland ("Entstehung, Charakter und Herkunft des 
sog. Westlichen Textes untersucht an die Apostelgeschichte," Ephemerides 
theologicae Lovanienses, lxii 11985], pp. 51-65) suggests that there were three 
stages in the development of this form (or of such forms) of text. In the 
second century, copyists introduced interpolations, omissions, and alter- 
ations in the text of Acts that tended in the direction of the Western type of 
text. In the first half of the third (?) century, a redactor revised a manuscript 
that contained a form of text belonging to the first stage, and this resulted in 
a text embodying the well-known Western characteristics. At the third stage, 
the redactor's exemplar was copied by various persons who dealt with the 
text in a rather free manner. 

16. The opinion of K. Aland ("Alter und Entstehung des D-Textes im 
Neuen Testament; Betrachtung zu P 69 und 0171," Miscelldnia papirologica 
Ramon Roca-Puig, ed. by Sebastia Janeras [Barcelona, 1987], pp. 37-61), 
that this form of text arose only in the second half of the third century, 
appears to be contradicted by the patristic and versional evidence. See, 
e.g., Bart D. Ehrman, "Heracleon and the 'Western' Textual Tradition," New 
Testament Studies, xl (1994), pp. 161-79. 

17. C. M. Martini, "La Tradition textuelle des Actes des Apotres et les 
tendences de l'eglise ancienne," in Les Actes des Apotres; Traditions, redac- 
tion, theologie, ed. by J. Kremer (Leuven, 1979), p. 34. On the distinction 
between original and secondary Western readings, see G. Zuntz's wide- 
ranging discussion "On the Western Text of the Acts of the Apostles" in 
Opuscula selecta (Manchester, 1972), pp. 189-215; J. Neville Birdsall, "The 
Western Text in the Second Century," in Gospel Traditions in the Second 
Century, ed. by William L. Petersen (Notre Dame, 1989), pp. 3-17. 

The Practice of New Testament Textual Criticism 309 

result of retranslation into Greek from Syriac or from the Old Latin, 
perhaps to make the Greek agree with the Syriac or the Old Latin in 
a bilingual manuscript. Since most of the striking elements of the 
Western text occur in Luke and Acts, several scholars theorized that 
Luke himself had issued two editions of these books, one being 
somewhat longer than the other, the longer edition being the West- 
ern text. (There is difference of opinion over whether the longer or 
shorter was the first edition.) 

The most important witnesses of the Westen text are Codex 
Bezae and the Old Latin manuscripts, 18 all of which are characterized 
by longer or shorter additions and by certain striking omissions. So- 
called Western texts of the Gospels, Acts, and Pauline Epistles circu- 
lated widely 19 not only in North Africa, Italy, and Gaul (which are 
geographically western) but also in Egypt 20 and (in somewhat differ- 
ent text forms) the East. These latter text forms are represented by 
the Sinaitic and Curetonian manuscripts of the Old Syriac, by many 
of the marginal notes in the Harclean Syriac, and perhaps by the 
Palestinian Syriac. 

Finally, it may be noted that Westcott and Hort regarded the 
Western text as almost totally corrupt and accepted as original in it 
only what they called "Western non-interpolations." As was men- 
tioned above, subsequent scholars (e.g., Adalbert Merx and A. C. 
Clark) reacted against this one-sided view with an equally one-sided 
preference for the Western text. Today, such extreme positions for 
and against Western forms of text find little favor, for most textual 
scholars recognize that all of the pre-Koine forms of text deserve a 
hearing and that any one of them may preserve original readings that 
have been lost to the other text types. 

Study of the Western text continues unabated. Although recent 
attempts to produce comprehensive theories on the Western group 
of texts have little in common, a comparison of their proposals, 
methods, and results may bring the riddle of the Western text a little 
closer to solution. Significant surveys of such attempts can be con- 
sulted in Joel Delobel's "Focus on the 'Western' Text in Recent Stud- 
ies," Ephemerides tbeologicae Lovanienses, lxxiii (1997), pp. 401-10, 

18. See Philip Burton, The Old Latin Gospels: A Study of'Iheir Texts 
and Language (Oxford, 2000). 

19. The Catholic Epistles and the Book of Revelation seem not to have 
existed in a characteristically Western form of text. 

20. See p. 277, n. 10. 

310 The Text of the New Testament 

where more than 40 books and articles are discussed, and in the 
volume Codex Bezae: Studies from thel.unel Colloquium, June 1994, 
edited by D. C. Parker and C.-B. Amphoux (Leiden, 1996), where the 
three areas of recurring interest of the score of contributors are the 
appropriateness of identifying Codex Bezae as part of the Western 
text, the possibility that Codex Bezae represents the original text, 
and the place of origin of the codex. 

Western Witnesses 

Gospels: K (in John 1.1-6.38), D, W (in Mark 1.1-5.30), 0171, the Old 

Latin, Syr s , Syr c (in part), early Greek and Latin fathers, Tatian's 

Acts: p 29 , p 38 , p 48 , D, 383, 614, Syr h m «, early Greek and Latin fathers, 

and the Commentary of Ephraem (preserved in Armenian). 
Pauline Epistles: the Greek-Latin bilinguals D p , E p , F p , and G p ; Greek 

fathers to the end of the third century; Old Latin and early Latin 

fathers; and Syrian fathers to about A.D. 450. 


B. H. Streeter identified the text that Origen used at Caesarea 
and associated it with the text in 0, fam. 1, fam. 13, and other wit- 
nesses. Subsequent investigations by Lake, Blake, and New showed 
that the Caesarean text probably originated in Egypt and was 
brought by Origen to Caesarea, whence it was carried to Jerusalem 
(a number of Caesarean witnesses contain the so-called Jerusalem 
colophon; see the description of Codex 157 in Chapter 2), to the 
Armenians (who had a colony in Jerusalem at a very early date), and 
then to the Georgians (Codex Koridethi belongs to Georgia). 

The special character of the Caesarean text is its distinctive 
mixture of Western and Alexandrian readings. According to Lagrange, 
its maker evidently knew both and made a kind of compromise; in 
substance he followed the Alexandrian text while retaining any 
Western readings that did not seem too improbable, for the latter text 
was widely current, although the former was the better. One may 
also observe a certain striving after elegance and, thus, consideration 
for the needs of the Church. 21 A similar view is advanced by Globe, 
who maintains that in their early form the "Caesarean variants 
resemble the conscious harmonizations, paraphrases and smoothing 

21. Lagrange, op. cit., pp. 163 ff. 

The Practice of New Testament Textual Criticism 311 

of grammatical details also found in Western sources." 22 Based on 
his statistical analysis, Hurtado comes to a similar conclusion, arguing 
that this type of text is "a form of Western text as it was shaped in the 
East." 23 

According to investigations made by Ayuso and others, 24 it is nec- 
essary to distinguish two stages in the development of the Caesarean 
text (at least for Mark). The Old Egyptian text that Origen brought 
with him to Caesarea may be called "pre-Caesarean." This is pre- 
served in p 45 , W (in Mark 5.31-16.20), fam. 1, fam. 13, 28, and many 
Greek lectionaries. At Caesarea and in its subsequent development, 
the pre-Caesarean text took on the form to which we are led back by 
the common evidence of <9, 565, and 700, many of the citations of 
Origen and Eusebius, the Old Armenian and Old Georgian versions 
(this form is the Caesarean text proper), and to some extent, the Old 
Syriac (Syr 5 ' c ). The most recent study by Amphoux also argues for 
two stages in the Caesarean text but maintains that the earliest is 
to be associated with Antioch rather than Caesarea. 25 Hurtado's 
study, while confirming the close relationship of W with p 4 ', shows 
that fam. 13 is a secondaiy witness to the W-p 45 text and that the three 
witnesses are not related to the Caesarean text represented by 565, 
700, and 0. Consequently, "the 'pre-Caesarean' witnesses are not 
Caesarean at all." 26 

22. See Alexander Globe, "Serapion of Thmuis as a Witness to the 
Gospel Text Used by Origen in Caesarea," Novum Testamentum, xxvi 
(1984), pp. 126 f. Globe concludes that Serapion's quotations demonstrate 
the use in fourth-century provincial Egypt of a Gospel text remarkably 
similar to the one used a century earlier in Caesarea by Origen. Indeed, in 
the passages available for comparison, Serapion's text is even closer to 
Origen's than that of Eusebius, the fourth-century father from Caesarea. 

23. Larry W. Hurtado, Text-Critical Methodology and the Pre-Caesarean 
Text: Codex Win the Gospel of Mark (Grand Rapids, MI, 1981), p. 88. 

24. For a survey of the history of the investigation of the Caesarean 
text of the Gospels, see Metzger, op. cit.. pp. 42-72. 

25. Christian-B. Amphoux, "Le texte evangelique de Cesaree et le type 
de texte 'Cesareen' des Evangiles," Filologia Neotestamentaria. xii (1999), 
pp. 3-16. Amphoux concludes, "Le type de texte 'Cesarean' ne se confond 
pas avec le texte lu a Cesaree: telle est la consequence des appellations geo- 
graphiques donnees aux types de texte grecs du Nouveau Testament." (For a 
more detailed summary, see New Testament Abstracts, slv [2001], pp. 232-3.) 

26. Hurtado, op. cit., pp. 88 f. 

312 Thf Text of hie New Testament 

In short, the Caesarean text appears to be the most mixed and 
the least homogeneous of any of the groups that can be classified as 
distinct text types. 


It is widely agreed that the Alexandrian text was prepared by 
skillful editors, trained in the scholarly traditions of Alexandria. 27 The 
text on which they relied must have already been ancient in all 
important points. For much of the late nineteenth and twentieth cen- 
turies, the two chief witnesses to this form of text were B and », dat- 
ing from about the middle of the fourth century. With the discovery, 
however, of p 66 and p 75 , both dating from about the end of the sec- 
ond or the beginning of the third century, proof became available 
that Ilort's Neutral text goes back to an archetype that must be put 
early in the second century. 28 This earlier form of the Alexandrian 
text, which may be called the "primary" Alexandrian text, is gener- 
ally shorter than the text presented in any of the other forms, the 
Western being the longest. Furthermore, the primary Alexandrian 
text appears not to have undergone the systematic grammatical and 
stylistic polishing that was given to other texts, including the later 
form of the Alexandrian text itself. 29 

Though most scholars have abandoned Hort's optimistic view 
that Codex Vaticanus (B) contains the original text almost unchanged 
except for slips of the pen, they are still inclined to regard the 
Alexandrian text as on the whole the best ancient recension and the 
one most nearly approximating the original. 

27. See G. Zuntz, The Text of the Epistles, a Disquisition upon the 
Corpus Paulinum (London, 1953), pp. 272-6. 

28. See C. L. Porter, "Papyrus Bodmer XV (p 75 ) and the Text of Codex 
Vaticanus," Journal of Biblical Literature, lxxxi (1962), pp. 363-76; Kurt 
Aland, "Neue neutestamentliche Papyri," New Testament Studies, xi 
(1964-5), pp. 5-21; New Testament Studies, xii (1965-6), pp. 195-210 (repr. 
with additions in Aland's Studien zur Uherlieferung des Neuen Testaments 
und seines Textes [Berlin, 1967], pp. 155-72); Carlo M. Martini, II problema 
della recensionalitd del codice B alia luce delpapiro Bodmer XIV (Rome, 
1966); Gordon D. Fee, "p 75 , p 66 , and Origen; The Myth of Early Textual 
Recension in Alexandria," in New Dimensions in New Testament Studies. 
ed. by Richard Longenecker and Merrill Tenney (Grand Rapids, MI, 1974), 
pp. 19-45. 

29. For the terminology of "primary" and "secondary" Alexandrian wit- 
nesses and a discussion of their internal relationships, see Bart D. Hhrman, 
Didymus the Blind and the Text of the Gospels (Atlanta, 1986), esp. pp. 262-7. 

The Practice of New Testament Textual Criticism 313 

Alexandrian Witnesseses 

Primary Alexandrian: p 45 (in Acts), p 46 , p 66 , p 75 , K (except for John 
1.1-8.38), B, Sah (in part), Clem, Orig, and most of the papyrus 
fragments with Pauline text. 

Secondary Alexandrian: (C), L, T, W (in Luke 1.1-8.12 and John), (X), 
Z, A (in Mark), S, ^(in Mark, partially in Luke and John), 33, 579, 
892, 1241, Boh, Didymus the Blind, and Athanasius 

Acts: p 50 , A, (C), V, 33, 81, 104, and 326. 

Pauline Epistles: A, (C), HP, I, W, 33, 81, 104, 326, and 1739. 

Catholic Epistles: p 20 , p 23 , A, (C), W, 33, 81, 104, 326, and 1739. 

Revelation: A, (C), 1006, 1611, 1854, 2053, 2344, and (less good) p 47 
and ». 

After having ascertained the text types represented by the evi- 
dence supporting each of the variant readings under examination, 
the student should draw a tentative conclusion as to the preferred 
reading on the basis of considerations bearing on the age of the 
manuscripts, the geographical spread of the witnesses that join in 
support of a given reading, and the textual type to which it belongs. 
Due appreciation of the implications of genealogical relationship 
among manuscripts prevents one from favoring a reading merely 
because a large number of witnesses may support it. 

2. Internal Evidence 

The next step in the process of evaluating variant readings is to ap- 
peal to internal evidence, beginning with transcriptional probabilities. 
Which reading is the more difficult, that is, more difficult to the scribe? 
Other things being equal, the reading that puzzled the scribe is most 
likely to be correct. But there is a point at which what is relatively dif- 
ficult becomes absolutely difficult and, therefore, impossible to be 
regarded as original. 

Some readings were favored by scribes because they supported 
current beliefs and practices in their part of the Christian world. 
Hence, the textual critic will need to have the fullest knowledge of 
the development of Christian doctrine and cult, as well as all the 
heretical aberrations in the early Church. Obviously, acquaintance 
with paleographical features of majuscule and minuscule hands, along 
with a knowledge of dialectical variations in Greek orthography and 
syntax, will often suggest the correct evaluation of a variant reading. 
When dealing with a passage in the Synoptic Gospels, it is necessary 
to examine the evidence of parallel passages. The harmonization of 

314 Tiik Text of ire Nkw Testament 

the Evangelists is by definition a secondary procedure; therefore, the 
supreme rule for editors of the text is to give each Gospel its own 
proper character. This means that ordinarily the reading that differs 
from a parallel passage (particularly when the evidence for the read- 
ing of the parallel is firm) should be preferred. Likewise, in quotations 
from the Old Testament, the text and apparatus of the Septuagint must 
be consulted. Since scribes tended to make New Testament quotations 
conform to the text of the Septuagint, readings that diverge from 
the Old Testament should not be rejected without the most careful 

Finally, the student may appeal to intrinsic probability. The read- 
ing deemed original should be in harmony with the author's style and 
usage elsewhere. Since, however, it is conceivable that several variant 
readings may fulfill this requirement, the textual critic should be 
guided more by negative judgments delivered by intrinsic evidence 
than by positive judgments. The appropriate question is whether in- 
trinsic evidence opposes the conclusion commended by genealogical 
considerations, the geographical distribution of witnesses, and tran- 
scriptional probabilities. 

Sometimes the only reading that seems to be in harmony with 
the author's usage elsewhere is supported by the poorest external 
evidence. In such cases, the decision of the textual critic will be 
made in accordance with a general philosophy of textual methodol- 
ogy. It is probably safest for the beginner to rely on the weight of 
external evidence rather than on what may be an imperfect knowl- 
edge of the author's usage. 

In the course of time, the student will observe that generally the 
reading that is supported by a combination of Alexandrian and West- 
ern witnesses is superior to any other reading. There is, however, an 
exception to this observation: in the Pauline Epistles, the combina- 
tion of B, D, and G is ordinarily not of great weight. The reason for 
this is that though R is purely Alexandrian in the Gospels, it has a 
certain Western element in the Pauline Epistles. Hence, the combina- 
tion of B plus one or more Western witnesses in Paul may mean only 
the addition of one Western witness to others of the same class. 30 

The combination of Western and Caesarean witnesses does not 
usually possess exceptional weight, for these two types of text are 
closely related, especially in the early periods. 

30. The same applies to Codex Sinaiticus in John 1.1-8.38, where it 
has a Western text. 

The Practice of New Testament Textual Criticism 315 

In the evaluation of readings that are supported by only one 
class of witnesses, the student will probably find that true readings 
survive frequently in the Alexandrian text alone, less frequently in 
the Western group alone, and very rarely only in Caesarean wit- 
nesses. As a rule of thumb, the beginner may ordinarily follow the 
Alexandrian text except in the case of readings contrary to the cri- 
teria that are responsible for its being given preference in general. 
Such a procedure, however, must not be allowed to degenerate into 
merely looking for the reading that is supported by B and s (or 
even by B alone, as Hort was unfairly accused of doing); in every 
instance, a full and careful evaluation is to be made of all the vari- 
ant readings in the light of both transcriptional and intrinsic proba- 
bilities. The possibility must always be kept open that the original 
reading has been preserved alone in any one group of manuscripts, 
even, in extremely rare instances, in the Koine or Byzantine text. 

It remains now to put into practice these principles. Lest, how- 
ever, the student imagine that the procedures of criticism are stereo- 
typed and doctrinaire, this section may be concluded on a lighter vein 
with a quotation from a scintillating essay on textual criticism by A. E. 

Textual criticism is not a branch of mathematics, nor indeed an exact 
science at all. It deals with a matter not rigid and constant, like lines and 
numbers, but fluid and variable; namely the frailties and aberrations of 
the human mind, and of its insubordinate servants, the human fingers. 
It is therefore not susceptible of hard-and-fast rules. It would be much 
easier if it were; and that is why people try to pretend that it is, or at 
least behave as if they thought so. Of course you can have hard-and- 
fast rules if you like, but then you will have false rules, and they will 
lead you wrong; because their simplicity will render them inapplicable 
to problems which are not simple, but complicated by the play of per- 
sonality. A textual critic engaged upon his business is not at all like 
Newton investigating the motions of the planets: he is much more like 
a dog hunting for fleas. If a dog hunted for fleas on mathematical prin- 
ciples, basing his researches on statistics of area and population, he 
would never catch a flea except by accident. They require to be treated 
as individuals; and every problem which presents itself to the textual 
critic must be regarded as possibly unique." 

31. A. E. Housman, "The Application of Thought to Textual Criticism" 
in Proceedings of the Classical Association, August, 1921, xviii (London, 
1922), pp. 68-9. (This essay is now more widely available in John Carter's 
edition of several of Housman's critical writings, entitled Selected Prose 
[Cambridge, 196l], pp. 131-50.) 

316 The Text of the New Testament 

III. The Textual Analysis of 
Selected Passages 32 

The following passages have been chosen in order to provide 
illustrative examples of various kinds of text-critical problem. To pre- 
vent monotony in the exposition and to emphasize that no one stereo- 
typed method of textual analysis is suited to all problems, the presen- 
tation of the kinds and nature of the evidence will be varied in 
sequence and in development of argument. The discussion begins 
with relatively simple problems, for which one can usually find clear 
and unambiguous solutions, and concludes with more complex prob- 
lems, where the probabilities are much more evenly divided and the 
critic must sometimes be content with choosing the least unsatisfactory 
reading or even admitting that there is no clear basis for choice at all. 

It is customary in a critical apparatus to use abbreviations of cer- 
tain Latin words as a concise and international working language. 
The following are in general use: 

pc (pauci) = a few other manuscripts 

al (alii) = other manuscripts 

pm (permultt) = very many other manuscripts 

pi (plerique) — most other manuscripts 

rell (reliqui) = the remaining witnesses 

vid (videtuf) = as it seems, apparently 

omn (omnes) = all manuscripts 

codd (codices) = manuscripts of a version or Church father as 

distinguished from the edition 
ap (apud) = in the writings of, on the authority of (e.g., 

Papias ap Eusebius) 
pt (partim) = divided evidence (e.g., Orig pt signifies that 

Origen is inconsistent in his quotations of 

the same passage) 
2/4 = divided evidence (e.g., Orig 2/4 signifies that 

in two cases out of four quotations of the same 
passage Origen supports a given reading) 

An asterisk placed after the siglum of a manuscript indicates that 
the manuscript at the passage referred to has been corrected and that 

32. See also Elizabeth G. Edwards, "On Using the Textual Apparatus 
of the UBS Greek New Testament," The Bible Translator, xxviii (Jan. 1977), 
pp. 121-42. 

The Practice of New Testament Textual Criticism 317 

the original reading is being cited; a superior letter (e.g., c ) placed 
after the siglum indicates that the corrected reading is being cited. 
Sometimes the work of more than one corrector can be differenti- 
ated (for the several ways in which such information is cited, see 
the descriptions of Codex Sinaiticus, pp. 62-7 and Codex Ephraemi, 
pp. 69-70). When the siglum of a manuscript is enclosed within 
parentheses, this signifies that the manuscript supports the chief 
point of the variant reading but differs in minor respects. 

In this connection, a warning may not be out of place: in some 
apparatus critici of the New Testament, the sigla of majuscule 
manuscripts are often cited without the superior letter (or the inferior 
numeral) that serves to distinguish certain manuscripts from others 
designated by the same siglum. Thus, D simpliciter often stands for 
Codex Claromontanus (instead of D p or D 2 ) as well as for Codex 
Bezae. In such instances, one must be alert to distinguish between 
the two manuscripts by observing whether the variant reading oc- 
curs in the Gospels or Acts (in this case, D = Codex Bezae) or 
whether the variant occurs in the Pauline Epistles (in this case, D = 
Codex Claromontanus). In the following textual analyses, the wit- 
ness's siglum with the superior letter (when this is appropriate) will 
be used in accordance with the descriptions of manuscripts given in 
Chapter 2: that is to say, a after a siglum indicates a manuscript that 
contains the Acts and the Catholic Epistles, p indicates the Pauline 
Epistles, and r indicates the Book of Revelation. 

In the King James Version of Acts 6.8, Stephen is described as "full 
of faith and power" (jrlijgrjg marewg xai 5vvdfA.eaj<;), whereas in the 
New Revised Standard Version, he is said to be "full of grace and 
power" (jrAijQqg xdgiTog xai dwdfiecog). The difference in the English 
versions represents not variant renderings of the same Greek word but 
variant readings in the basic Greek text. The textual evidence, which 
involves four variant readings, is as follows: 

1. "Grace" (^dgurog) is read by p 8 45vid 74 , s, A, B, D, more than 20 
minuscule manuscripts, Vulg, Sah, Boh, Syr p , Arm, and Eth (the 
last reads %a.Qixog deov). 

2. "Faith" (moreajg) is read by H a , P a , many minuscule manuscripts, 
Syr 11 , Gregory of Nyssa, and Chr. 

3. "Grace and faith" (xagirog xai moreajg) is read by E a . 

4. "Faith and grace of the Spirit" {moxewg xai %aoiTog nvevfiarog) is 
read by W. 

318 The Text of the New Testament 

Of these four variant readings, it is obvious that either the first 
two are independent abridgements of the longer readings or the third 
and fourth readings arose from combining the elements of the first 
two. Considerations of both external evidence and internal probabil- 
ity unite to demonstrate that readings 3 and 4 are secondary, being 
alternative conflations of the other two. Reading 3 is supported by 
the majuscule manuscript E a , which dates from the sixth century and 
is one of the earliest representatives of the Koine or Byzantine type 
of text in Acts. Reading 4 is supported by the majuscule manuscript 
W, which dates from the eighth or ninth century and has a mixed 
type of text in Acts. Transcriptional considerations lead one to con- 
clude that both readings 3 and 4 presuppose the priority of the other 
two readings, for it is easier to believe that a scribe, knowing the 
existence of readings 1 and 2, decided to join them, lest the copy that 
he was writing lose one or the other, than to believe that two scribes 
independently took offense at the longer reading and that each 
chose to perpetuate half of it in his copy. Thus, external evidence, 
which is meager in extent and relatively late in date, and transcrip- 
tional probabilities unite against the originality of readings 3 and 4. 

Variant reading 2 is supported by two majuscule manuscripts, H a 
of the ninth century and P a of the tenth century, both representative 
of the Koine or Byzantine type of text. The majority of the minuscule 
manuscripts join these two majuscule witnesses. The earliest witness 
is Gregory of Nyssa, who died A.D. 395. 

Variant reading 1 is supported by a wide variety of witnesses, in- 
cluding representatives of the major pre-Koine types of text. Codices 
Sinaiticus and Vaticanus, both of the fourth century, are the earliest 
and best majuscule representatives in Acts of the Alexandrian type of 
text. Codex Bezae, of the fifth century, is the chief Greek represen- 
tative of the Western group of witnesses. Codex Alexandrinus, of the 
fifth century, and p 74 , dating from about the seventh century, have a 
mixed type of text. The evidence of the early versions, including the 
Latin, Syriac, Coptic, and Armenian, reflects the wide geographical 
area over which the reading was accepted. The external evidence in 
support of reading 1 is, therefore, far superior in point of age and di- 
versity of text type to that supporting reading 2. 

Internal probabilities likewise favor reading 1. If the account 
originally stated that Stephen was "full of faith," there is no dis- 
cernible reason why a scribe should alter it to "full of grace." On the 
other hand, in view of the statement made three verses earlier that 
Stephen was a man "full of faith and the Holy Spirit" (verse 5), it is 

The Practice of New Testament Textual Criticism 319 

easy to understand that in transcribing the later statement in verse 8 
copyists would be likely, either consciously or unconsciously, to 
substitute moTEoq, which they recalled from the earlier passage, for 
the correct reading xagtrog. The presence of jrvevfiarog in reading 
4 is to be explained in the same way. 

Thus, the converging of several strands of evidence, both exter- 
nal and internal, leads one to the firm conclusion that the author of 
Acts 6.8 wrote ithrfgr/g ^dptrog xai dwd^ewg. 

Not a few New Testament manuscripts incorporate here and 
there interesting details, some of which may be historically correct. 
The story of the woman taken in adultery, for example, has many 
earmarks of historical veracity; no ascetically minded monk would 
have invented a narrative that closes with what seems to be only a 
mild rebuke on Jesus' part: "Neither do I condemn you; go, and do 
not sin again." At the same time, the pericope, which is usually 
printed as John 7.53-8.11, must be judged to be an intrusion into the 
fourth Gospel. 33 

The account is lacking in the best Greek manuscripts: it is absent 
from p 66 , p 75 , «, 13, L, N, T, W, X, A, 0, V, 33, 157, 565, 1241, and fam. 
1424. Codices A and C are defective at this point, but it is highly 
probable that neither contained the section as there would not have 
been space enough on the missing leaves to include it along with 
the rest of the text. The Old Syriac (Syr s,c ) and the Arabic form of 
Tatian's Diatessaron betray no knowledge of the passage, nor is it 
contained in the best manuscripts of the Peshitta. Likewise, the old 
Coptic Churches did not include it in their Bible, for the Sahidic, the 
sub-Achmimic, and the older Bohairic manuscripts lack it. Some 
Armenian manuscripts as well as the Old Georgian version omit it. In 
the West, the passage is absent from the Gothic version and from 
several Old Latin manuscripts {a, f, /* q). 

Even more significant is the fact that no Greek Church father for 
1,000 years after Christ refers to the pericope as belonging to the 
fourth Gospel, including even those who, like Origen, Chrysostom, 
and Nonnus (in his metrical paraphrase), dealt with the entire 
Gospel verse by verse. Euthymius Zigabenus, who lived in the first 
part of the twelfth century, is the first Greek writer to comment on 

33. A significant study is still Ulrich Becker, Jesus und die Ehebrecherin 
(Berlin, 1963). 

320 Thk Text of the New Testament 

the passage; and even he declares that the accurate copies of the 
Gospel do not contain it. 34 

When one adds to this impressive and diversified list of external 
evidence the consideration that the style and vocabulary of the peri- 
cope differ markedly from the rest of the fourth Gospel and that it 
interrupts the sequence of 7.52 and 8.12 f., the case against its being 
of Johannine authorship appears to be conclusive. 

The earliest Greek manuscript known to contain the passage is 
Codex Bezae, of the fifth century, which is joined by several Old Latin 
manuscripts (aur, c, e,f,f 2 ,j, r l ). The pericope is obviously a piece 
of floating tradition that circulated in certain parts of the Western 
Church. 35 It was subsequently inserted into various manuscripts at 
various places. Most scribes thought that it would interrupt John's 
narrative least if it were inserted after 7.52 (D, E, F, G, H, K, M, S, U, 
r, A, 77, 28, 579, 700, and 1579). Others placed it after 7.36 (MS. 225) 
or after 21.24 (fam. 1, 1076, 1570, 1582). The revision of the Old 
Georgian version, made in the eleventh century by George the 
Athonite, contains the passage after 7.44. The scribe of an ancestor 
of fam. 13 inserted it in another Gospel altogether, after Luke 21.38. 
Significantly, in many of the manuscripts that contain the passage, it 
is marked with an obelus (S) or an asterisk (E, M, A), indicating that, 
though the scribes of these manuscripts included the account, they 
were aware that it lacked satisfactory credentials. 

A few of the manuscripts that report the incident also include an 
interesting expansion at the close of 8.8. More than one reader of the 
statement that Jesus "bent down and wrote with his finger on the 
ground" must have wondered what it was that the Lord wrote. An 
unknown copyist satisfied this natural curiosity by adding the words 
"the sins of each of them." 36 

The best disposition to make of the pericope as a whole is 
doubtless to print it at the close of the fourth Gospel, with a footnote 

34. The fourth-century Didymus the Blind, however, refers to it, or a 
similar story, as found "in some Gospels." For a full study, with reference to 
other early patristic sources, see Bart D. Ehrman, "Jesus and the Adulteress," 
New Testament Studies, xxxiv (1988), pp. 24-44. 

35. For a hypothetical reconstruction of different forms of the story 
known in different parts of the church, see ibid., pp. 34-8. 

36. See David C. Voss, "The Sins of Each One of Them," Anglican 
Theological Review, xv (1933), pp. 321-3. 

The Practice of New Testament Textual Criticism 32 1 

advising the reader that the text has no fixed place in the ancient 

witnesses. 37 

An interesting example of a comment that was felt to be open to 
misinterpretation is found in John 7.37-9: 

On the last day of the festival, the great day. while Jesus was standing 
there, he cried out, "Let anyone who is thirsty come to me and let the 
one who believes in me drink. As the scripture has said, 'Out of the be- 
liever's heart shall flow rivers of living water'." Now he said this about 
the Spirit, which believers in him were to receive; for as yet there was 
no Spirit, because Jesus was not yet glorified. 

In the final sentence, the clause "for as yet there was no Spirit" 
(ovjtco yag r\v Jivevfia) appears in seven different forms: 

1. Jivevfia: p 66c , p 75 , s, N\ T, 0, 17, W, 1079, 1546, Cop bo ( ?°, Arm, 
Geo, Orig. 8r 

2. icvevfia ayiov. j> 66 *, L, W, X, r, A, A, fam. 1, 13, 28, 33, 565, 579, 
700, 892, Byz. 

3. nvev/ua ayiov en avroig-. D, f, Goth. 

4. Jivevfia dedofievov-. a, aur, b, c, ff 2 , I, r 1 , Vulg, Syr 5, Ci p , Eus, 
Jerome, Aug. 

5. nvevpa ayiov dedofiEvov. B, 053, 1230, e, q, Syr pa1, h . 

6. "for they had not yet received [the] Spirit": Cop sah ' subach . 

7. "for the Holy Spirit had not yet come": Eth. 

A little reflection will make it obvious that the reading that ex- 
plains the rise of all the others is reading 1, Jivevfia. Many scribes 
were doubtless perplexed by the bare and ambiguous statement "for 
there was no Spirit, because Jesus was not yet glorified." Lest this be 
taken to affirm that the Spirit was not in existence prior to Jesus' glo- 
rification, modifications were introduced to relieve the difficulty. 
Several Western witnesses (D, f, Goth) read 3, "for the Holy Spirit 
was not yet in them." Other witnesses add the verb "given" (as in 
readings 4 and 5), "received" (reading 6), or "come" (reading 7). 

The adjective ayiov (readings 2, 3, and 5) is a most natural kind 
of addition, which many scribes could make independently of one 

37. For a discussion of the origin of the pericope and the problem of 
why it was so late in being incorporated into the canonical text, see Harald 
Riesenfeld, "The Pericope de adultera in the Early Christian Tradition," in 
The Gospel Tradition (Philadelphia, 1970), pp. 95-110. 

322 The Text of the New Testament 

another. (The correction found in p 66 , deleting ayiov, is in keeping 
with the observed vigilance of this scribe in correcting his own 
inadvertent errors.) In this case, Codex Vaticanus is doubly in error 
(reading 5), having added both ayiov and a predicate verb. 

The evidence for reading 2 can be joined to that of 1 with re- 
spect to resisting the temptation to add a predicate verb. There is 
thus a very widespread and diversified constellation of witnesses in 
support of the more difficult and shorter reading. It can scarcely be 
doubted, therefore, that the original text had simply ovjtu) yag rjv 
TTvevfia, 58 

How did Mark end his Gospel? 39 Unfortunately, we do not 
know; the most that can be said is that four different endings are cur- 
rent among the manuscripts but probably none of them represents 
what Mark originally intended. These four endings may be called the 
short ending, the intermediate ending, the long ending, and the long 
ending expanded. The evidence for each of them is as follows: 

1. The last 12 verses of Mark (16.9-20) are lacking in the two earli- 
est parchment codices, B and K, in the Old Latin manuscript k, the 
Sinaitic Syriac, many manuscripts of the Old Armenian version, 
the Adysh and Opiza manuscripts of the Old Georgian version, 
and a number of manuscripts of the Ethiopic version. Clement of 
Alexandria, Origen, and Ammonius show no knowledge of the 
existence of these verses; other Church fathers state that the sec- 
tion is absent from Greek copies of Mark known to them (e.g., 
Jerome, Epist. cxx.3, To Hebidia, "Almost all the Greek copies 
do not have this concluding portion"). The original form of the 
Eusebian sections makes no provision for numbering sections 
after 16.8. Not a few manuscripts that contain the passage have 
scholia stating that older Greek copies lack it (e.g., MSS. 1, 20, 22), 

38. This is the text lying behind the King James Version, the Revised 
Version of 1881, and the American Standard Version of 1901, all of which 
use italics to show what the translators have added for the sake of English 
readers: ". . . was not yet given." Since neither the New English Bible nor 
the New Revised Standard Version uses italics in this way, the inclusion of 
the verb "given" in these two versions is to be accounted for either as the 
result of license in translation or as the choice of what appears to be a 
secondary variant reading as the basic text. 

39. Among recent studies, see Paul L. Danove, The End of Mark's 
Story: A Methodological Study (Leiden, 1993), pp. 120-5; David C. Parker, 
The Living Text of the Gospels (Cambridge, 1998), pp. 124-47. 

The Practice of New Testament Textual Criticism 323 

and in other witnesses the passage is marked with asterisks or 
obeli, the conventional sigla used by scribes to indicate a spurious 
addition to a literary document. 

2. The intermediate ending ("But they reported briefly to Peter and 
those with him all that they had been told. And after this Jesus 
himself sent out by means of them, from east to west, the sacred 
and imperishable proclamation of eternal salvation") is present in 
several majuscule manuscripts of the seventh, eighth, and ninth 
centuries (L, V, 099, 0112) as well as in a few minuscule 
manuscripts (274 mg [Fig. 25], 579) and several ancient versions 
(k, Syr 11 m «, CopP', Eth codd ). 40 

3. The long ending, so familiar through the King James Version and 
other translations of the Textus Receptus, is present in the vast 
number of witnesses (including several that also contain the inter- 
mediate ending), namely, A, C, D, L, W, 0, most of the later 
majuscules, the great majority of the minuscules, and most of the 
Old Latin witnesses, such as Vulg, Syr c ' p , and Cop pl . It is probable 
that Justin Martyr, at the middle of the second century, knew 
this ending; in any case, Tatian, his disciple, included it in his 

4. The long ending in an expanded form existed, so Jerome tells us, 
in Greek copies current in his day; and since the discovery of W 
earlier this century, we now have the Greek text of this expansion 
(for a translation of the addition after verse 14, see p. 81). 

None of these four endings commends itself as original. The 
obvious and pervasive apocryphal flavor of the expansion as well as 
the extremely limited basis of evidence supporting it condemn it as 
a totally secondary accretion. 

The long ending, though present in a variety of witnesses, some 
of them ancient, must also be judged by internal evidence to be 
secondary. 41 For example, the presence of 17 non-Marcan words or 

40. For detailed evidence of the Coptic versions, see P. E. Kahle, 
Journal of Theological Studies, n.s., ii (1951), pp. 49-57. 

41. The late William R. Farmer attempted to account for the absence 
of the last 12 verses of Mark in certain early witnesses on the supposition 
that, during the second century, Alexandrian scribes, being uneasy with 
the references in verse 18 to "picking up snakes" and "drinking poison," 
decided to adopt techniques of Alexandrian Homeric scholars and conse- 
quently deleted all 12 verses in certain Greek manuscripts {The Last Twelve 
Verses of Mark [Society for New Testament Studies: Monograph Series, xxv; 

324 Tiik Text of the New Testament 

Kay iur»»vA.«A(C«"-rr%^6* (itim* <r»iV> 

„,* J -rep |"«f '*> «rkV"° •V*^**J"*»« K •w^tii'^k MAXff^LtWC 
r» o p fro •*(( «™u eu -lA-Usyv/f **£y «m » «• |*ir*»'u 

«*4 fci* fr <#Gau «frU v-ktblo -rfCtr I«j-rrl q4«ur«^ 

tbVjJI y *• *«4» 6p at' -*Crfy &f <m j»« p4** 

-*}•?•>•»■»{* {CM a»y <* A* -*v» tIs*m«- 
^ «trl«-T»4* *Xnro»}« tuM«4<>Jkf>« buotf •***»t» , 
jj' ' • "«H-»nV -**Wm f i*i» ( >i» «Wr»y Afivfif r&f- 

gg lUMTAM'TAtuf m rnxuiMCttit n T»m 

Figure 25 Greek Gospel MS. 274 (tenth century), Bibliotheque Nationale, 
Paris; Mark 16.6-20 and the intermediate ending of Mark (see §2 on p. 323). 
Actual size 9% X 6'A inches. 

The Practice of New Testament Textual Criticism 325 

words used in a non-Marcan sense, the lack of a smooth juncture 
between verses 8 and 9 (the subject in 8 is the women, whereas Jesus 
is the presumed subject in 9), and the way in which Mary is identified 
in verse 9 even though she has been mentioned previously (verse 1) 
indicate that the section was added by someone who knew a form of 
Mark that ended abruptly with verse 8 and who wished to provide 
a more appropriate conclusion. An Armenian manuscript of the 
Gospels, copied A.D. 989 (see Fig. 21, p. 118), contains a brief rubric 
of two words in the space at the end of the last line of verse 8 and 
before the last 12 verses: Ariston eritsou ("of the Presbyter Ariston"). 
Many have interpreted this as a reference to Aristion, a contemporary 
of Papias in the early second century and traditionally a disciple of 
John the Apostle; but the probability that an Armenian rubricator 
would have access to historically valuable tradition on this point is 
almost nil, especially if, as has been argued, the rubric was added in 
the thirteenth or fourteenth century. 42 

The internal evidence of the so-called intermediate ending is 
decidedly against its being genuine. Besides containing a high per- 
centage of non-Marcan words, its rhetorical tone differs totally from 
the simple style of Mark's Gospel. The mouth-filling phrase at the 
close ("the sacred and imperishable message of eternal salvation") 
betrays the hand of a later Greek theologian. 

Thus, we are left with the short ending, witnessed by the earli- 
est Greek, versional, and patristic evidence. Both external and inter- 
nal considerations lead one to conclude that the original text of the 
second Gospel, as known today, closes at 16.8. But did Mark intend 
to conclude his Gospel with the melancholy statement that the 
women were afraid (i<f>o/3ovvTO ydg)? Despite the arguments that 

Cambridge, 19741). But this explanation is unsatisfactory: (1) Farmer misun- 
derstood the work of Alexandrian philologists, who did not ordinarily 
delete lines from the Homeric epics but merely athetized with an obolos 
the line or lines deemed to be inauthentic; (2) no convincing explanation is 
given for why all 12 verses should have been deleted rather than only the 
two phrases in verse 18 that Farmer thought were embarrassing to mem- 
bers of the Alexandrian Church; and (3) it is entirely unrealistic to suggest 
that the abbreviated Alexandrian text could account for the short text in 
such widely scattered versions as those preserved in the earliest Latin, 
Syriac, Sahidic, Armenian, and Georgian manuscripts. Not by the wildest 
stretch of one's imagination can it be argued that all these reflect the 
influence of the hypothetical suppression of the passage in Alexandria. 
42. So Joseph Schafers in Biblische Zeitschrift, xiii (1915), pp. 24-5. 

326 The Text oe the; New Testament 

several modern scholars have urged in support of such a view, 43 it is 
difficult to believe that the note of fear would have been regarded as 
an appropriate conclusion to an account of the Evangel, or Good 
News. Furthermore, from a stylistic point of view, to terminate a 
Greek sentence with the word ydg is most unusual and exceedingly 
rare: only a relatively few examples have been found throughout the 
vast range of Greek literary works, and no instance has been found 
where ydg stands at the end of a book. Moreover, it is possible that 
in verse 8 Mark uses the verb E<pofiovvxo to mean "they were afraid 
of (as he does in four of the other occurrences of this verb in his 
Gospel). In that case, obviously something is needed to finish the 

It appears, therefore, that i(poj3ovvTO ydg of Mark 16.8 does not 
represent what Mark intended to stand at the end of his Gospel. 
Whether he was interrupted while writing and subsequently pre- 
vented (perhaps by death) from finishing his literary work or 
whether the last leaf of the original copy was accidentally lost before 
other copies had been made, 44 we do not know. All that is known is 
that more than one person in the early Church sensed that the 
Gospel is a torso and tried in various ways to provide a more or less 
appropriate conclusion. 45 

43. Eg. J. M. Creed, Journal of Theological Studies, xxxi (1932), 
pp. 175-80; Ernst Lohmeyer, Das Evangelium des Markus (Gottingen, 
1937), pp. 356-60; N. B. Stonehouse, The Witness of Matthew and Mark to 
Christ (Philadelphia, 1944), pp. 86-118; A. M. Parrer, The Class of Vision 
(London, 1948), pp. 136-46; R. H. Lightfoot, The Gospel Message of St . Mark 
(Oxford, 1950), pp. 80-97, 106-16. For a recent discussion of attempts to 
understand the literary significance of the short ending, see J. F. Williams, 
"Literary Approaches to the End of Mark's Gospel," Journal of the 
Evangelical Theological Society, xlii (1999), pp. 21-35. 

44. For a recent argument that the ending of Mark's Gospel (as well 
as its beginning!) was lost, see X. Clayton Croy, The Mutilation of Mark's 
Gospel (Nashville, 2003). 

45. Almost all textual studies and critical commentaries on the Gospel 
according to Mark agree that the last 12 verses cannot be regarded as 
Marcan; typical is the monograph by Clarence R. Williams, Ihe Appendices 
to the Gospel According to Mark: A Study in Textual Transmission ( Trans- 
actions of the Connecticut Academy of Arts and Sciences, vol. xviii; New 
Haven, 1915). 

In the nineteenth century, two major attempts were made to defend 
the genuineness of the long ending: John W. Burgon's The Last Twelve 

The Practice of New Testament Textual Criticism 327 

Text-critical analysis of the endings of Mark's Gospel has an im- 
portant bearing on the historical and literary source criticism of the 
Gospels. Since Mark was not responsible for the composition of the 
last 12 verses of the generally current form of his Gospel and since 
they undoubtedly were attached to the Gospel before the Church 
recognized the fourfold Gospels as canonical, it follows that the New 
Testament contains not four but five evangelic accounts of events 
subsequent to the Resurrection of Christ. 

In contrast with the length of the preceding textual variant, em- 
bracing 12 verses, one or two variant readings will now be examined 
that involve the presence or absence of a single letter. 

The Greek text that lies behind the traditional words of the an- 
gelic chorus at the birth of Jesus ("Glory to God in the highest, and on 
earth peace, good will toward men," Luke 2.14, King James Version) 
differs by only the letter sigma from the Greek text that lies behind 
the same verse in the New Revised Standard Version ("Glory to God 
in the highest, and on earth peace among those whom he favors"). In 
the former case, the King James translators followed the Textus 
Receptus, which reads evdoxia, supported by R, F, G, H, L, S, U, V, 
0, 3, Q, many minuscules, the Syriac, the Bohairic, the Georgian, 
Tatian, and Eusebius. In the latter case, the revisers followed a Greek 
text that reads the genitive case, evdoxiaq, supported by B*, k", A, D, 
W, 28, the Old Latin, Jerome's Vulgate, the Gothic, the Sahidic, 
Irenaeus, Cyril of Jerusalem, and the Latin fathers. 

The earliest Greek manuscripts of both the Alexandrian and 
Western groups are joined by significant versional and patristic evi- 
dence in support of evdoxtag, proving that this reading was widely 
disseminated in the West and known in Upper Egypt and Palestine 
as well. On the other hand, the nominative form of the word was 
also widely read, being current in Syria, in Egypt around the delta, 

Verses of the Gospel According to St. Mark Vindicated (London, 1871; repr. 
1959), and J. P. P. Martin's Introduction a la critique textuelle du Nouveau 
Testament; Partie pratique, vol. ii (Paris, 1884). Another attempt to rehabili- 
tate this position was made by M. van der Valk, "Observations on Mark 16, 
9-20 in Relation to St. Mark's Gospel," Humanitas, n.s., vi-vii (1958), 
pp. 52-95. This essay is a singular exhibition of how not to practise textual 
criticism! Omitting entirely all consideration of the external evidence of 
manuscripts, van der Valk concentrates on more or less irrelevant and 
speculative considerations in order to arrive at what appears to be a 
predetermined conclusion. See also n. 41. 

328 The Text of toe New Testament 

and at Caesarea (0, Eusebius) and dating from as early as the sec- 
ond century (Tatian). On the whole, the external evidence appears to 
favor the genitive case, for the combination of B, «, and W with D and 
the Old Latin outweighs the external support for the nominative case. 
Internal considerations confirm this judgment, for in the context the 
genitive case is the more difficult to construe and therefore would be 
more likely to be altered to the nominative than vice versa. Further- 
more, consideration of intrinsic suitability corroborates the other evi- 
dence, for the expression "men of Ihis, i.e., God's] good will lor favor]" 
is a perfectly good Semitic construction, appearing several times in 
the Hebrew hymns discovered at Qumran, 46 and thus is entirely con- 
gruous with the Semitic cast of the first two chapters of Luke. 

Luke 2.14, therefore, involves a twofold strophe ("glory . . . 
peace"), and not a threefold strophe ("glory . . . peace . . . good 
will," all in the nominative). The sense is that the birth of the Messiah, 
the Lord (verse 11), is the occasion for the ascription of glory to God 
in highest heaven and the enjoyment of peace on earth among peo- 
ple of God's good will, that is, those persons on whom his favor rests, 
chosen to be the recipients of the gift of the Messiah. 

Another interesting variant reading that arises from the presence 
or absence of a single Greek letter is found in 1 Thess. 2.7. Here, the 
reading is either "we were gentle (rjTtioi) among you, like a nurse 
tenderly caring for her own children" or "we were infants (vr/moi) 
among you, like a nurse tenderly caring for her own children." The 
word tfmoi is supported by a c , A, C b , DP (c) , K p , LP, ?p, 33, SyrP' h , Sah, 
Arm, Clem, Orig 2/3, Chr, and Theodore of Mopsuestia; the word 
vrjmoi is supported by p 65 , **, B, C, D p *, F p , GP, I, Old Latin, Vulg, 
Boh, Eth, Orig lat , Eph, Cyril, Jerome, Aug, and Ambrst. 

It is easy to see how each of these variant readings may have 
arisen, for the word that precedes the variant is Eyevrjdrj/uev. When 
the text was copied by dictation, the pronunciation of iyEvrfdrj/uev 
rjTiLoi would be almost indistinguishable from that of eyEvrjOrjuev 

46. See Claus-Hunno Hunzinger, "Neues Licht auf Lc. 2. 14," 
Zeitscbrift filr die neutestamentlicbe Wissenscbaft, xliv (1952-3), pp. 85-90; 
idem, "Ein weiterer Beleg zu Lc. 2. 14 &v0qu>jioi svdomag,'' Zeitscbrift fur 
die neutestamentlicbe Wissenscbaft, xlix (1958), pp. 129-31; cf. Ernest 
Vogt, "'Peace among Men of God's Good Pleasure' Lk. 2. 14," in The 
Scrolls and the New Testament, ed. by K. Stendahl (New York, 1957), 
pp. 114-17. 

The Practice of New Testament Textual Criticism 329 

vijmoi and vice versa. Likewise, when the words stand in majuscule 
script, ErENH0HMENNHniOI is very like ErENHOHMENH- 
niOI and vice versa. Furthermore, it must be remembered that TV at 
the end of a line was frequently indicated merely by a stroke above 
the preceding letter, thus ErENH0HME 

It is less easy to decide in which direction the change went. On 
the one hand, the weight and diversity of external evidence are 
clearly in favor of vqmoi, which is supported by the earliest form of 
the Alexandrian text (p 65 [third century], »*, and B), the Western text 
(D* and Old Latin), as well as a wide variety of versions and fathers. 
Such a constellation of witnesses led Lachmann, Westcott and Hort, 
Zimmer, 47 and Bover to print vijmoi in their text, a choice favored 
also by such commentators as Lightfoot, Findlay (doubtfully), 
Wohlenberg, Frame, and Milligan (doubtfully). On the other hand, 
Paul's violent transition in the same sentence from a reference to 
himself as an infant to the thought of his serving as a mother-nurse 
has seemed to most editors and commentators to be little short of 
absurdity; therefore, Tregelles, Tischendorf, Alford, Weiss, von Soden, 
Merk, and Vogels follow the Textus Receptus and print rjmoi, a 
reading that is supported by Bornemann, von Dobschiitz, Moffatt, 
Dibelius, Lemonnyer, Voste, Neil, and Rigaux and was adopted by 
the translators of the Revised Standard Version, the New English 
Bible, and the New Revised Standard Version. 

Those who are impressed by the external evidence supporting 
vrjmoi attempt to alleviate the difficulty in sense that this word in- 
troduces by pointing to the earlier context (verses 3-6), where Paul 
speaks not of his gentleness but of his unselfish love. He defends 
himself against the charge, not of being harsh, but of using flattery to 
gain an unfair advantage. He suggests his lack of guile and childlike 
innocence, so it is argued, by calling himself an "infant." The word 
for "infant" by association brings with it the idea of a "nurse," and 
with his characteristic rapidity of thought, Paul inverts the metaphor 
and now refers to himself as a "mother-nurse." Such an inversion of 
metaphor — a Christian teacher being first compared to the child and 
then to the mother — is quite in the apostle's manner; for example, in 
Gal. 4.19, the sudden shift in metaphor is even more startling ("My 

47. Besides his monograph Der Text der Thessakmicherbriefe 
(Quedlinburg, 1893), Friedrich Zimmer supported his view in his contribu- 
tion to Bernhard Weiss 's Festschrift Theologische Studien (Gottingen, 1897), 
pp. 248-73, esp. pp. 264-9. 

330 The Text of the New Testament 

little children, with whom I am again in travail until Christ be formed 
in you"). It is also significant that 12 of the 16 cases of vrjmog in the 
New Testament are found in Paul (including Eph. 4.14), whereas 
rjmog appears elsewhere in the Greek Bible only in 2 Tim. 2.24. 

Not all of these arguments in favor of vrjmoi, however, are as sig- 
nificant as they may seem to be at first sight. Whatever may be said of 
the more remote context, in the immediate context "gentle" is certainly 
the appropriate antithesis to what Paul has just disclaimed in verse 6b, 
namely, the assertion of his apostolic dignity or authority, while in 
connection with the following clauses it is immeasurably preferable 
to "infants." Furthermore, though it is true that Paul uses vrjmog fre- 
quently, it must not be overlooked that he never refers to himself as a 
babe but always applies it to his converts. Moreover, it is certainly not 
without importance that at 2 Tim. 2.24, more than one scribe suc- 
cumbed to the temptation to substitute the more familiar word vrjmog 
for the true text, rjmog (supported by DP*, E p *, F p , G p , and Eth). 48 

By way of striking a balance in the arguments pro et contra, it 
appears that here internal considerations should be allowed to take 
precedence over external evidence. Since vrjmoi, which is by far 
the more common word in the New Testament, could so easily have 
been introduced by carrying over the final v of the preceding word 
and the violence done to the sense when vrjmoi is read appears to 
be intolerable, one is entitled to apply Daniel Mace's crisp dictum 
that no manuscript is so old as common sense and to make a tenta- 
tive decision in favor of rjmoi. 49 

A knowledge of Hebrew and especially of Aramaic will occa- 
sionally throw light upon a variant reading in the Gospels. For 
example, the words of Jesus in Mark 14.25, "Truly I tell you, I will 
never again drink of the fruit of the vine until that day when I drink 

48. The opposite error of writing rjmog for vrjmog also occurs sporadi- 
cally in individual manuscripts, e.g., in A at Eph. 4.14 and in 33 at Heb. 5.33. 

49- For a discussion of the resulting metaphor, see Abraham Malherbe, 
"'Gentle as a Nurse': The Cynic Background to 1 Thess ii," Novum 
Testamentum, xii (1970), pp. 203-17. For arguments in favor of the other 
reading, see Beverly Gaventa, "Apostles as Babes and Nurses in 1 
Thessalonians 2:7," in Faith and History: Essays in Honor of Paul W. Meyer, 
ed. by John T. Carroll, Charles H. Cosgrove, and E. Elizabeth Johnson 
(Atlanta, 1990), pp. 193-207; and Jeffrey Weima, "'But We Became Infants 
Among You': The Case for NHIIIOI in 1 Thess 2.7," New Testament Stud- 
ies, xlvi (2000), pp. 547-64. 

The Practice of New Testament Textual Criticism 331 

it new in the kingdom of God," are transmitted in three different 
forms. The text that lies behind the English versions is (1) ovxen ov 
firj mco, supported by A, B, A, fam. 1, 13, most minuscules, aur, b, 
ff 2 , i, I, q, Vulg, Syr 5, p - h , Sah, Geo, and (Arm); there are, however, 
two other readings, (2) ov fir/ moo, supported by R, C, L, W, 47, W, 
1, 892, 1342, c, k, Vulg (1 MS.), Boh, Eth, and Jerome, and (3) ov jur/ 
JiQoaOd) melv (nelv), supported by D, 0, 565, a, f, and Arm. (0 
reads ovxen ov /Atj JigoaOoofiev melv) 

The third reading, found in Western and Caesarean witnesses, 
involves a Semitic idiom, meaning literally "I shall not add to 
drink. ..." This idiom appears frequently in the Septuagint, which 
has more than 100 examples of JigooxtOevai (jioooriOeodai) for the 
Hebraic rppin (with an infinitive), meaning TiaXiv ("again"). It ap- 
pears that the Eucharistic words of Jesus, which were undoubtedly 
spoken in Aramaic or Hebrew to the apostles, have been preserved 
in literalistic fashion in the third variant reading, whereas the other 
two readings provide alternative interpretations of the meaning, 
expressed in more idiomatic Greek. (The second reading, in fact, can 
be called a misinterpretation, for it omits the idea expressed by 
"again.") Whether the idiom that is preserved in the third variant read- 
ing is to be considered original to Mark's Gospel (as Wellhausen 50 
suggested), whether it is a secondary biblicism in imitation of the Sep- 
tuagint (as Schiirmann 51 argues), or whether it discloses influence 
from extracanonical reports of Jesus' Eucharistic words (as Jeremias, 52 
following Black, 53 prefers), it is in any case a most noteworthy variant 
reading that is eminently suitable in the context. 

Many scholars 54 have given attention to the textual problem found 
in Acts 20.28, the report of Paul's farewell address to the Ephesian 

50. Julius Wellhausen, Einleitung in die drei ersten Kvangelien, 2te 
Aufl. (Berlin, 1911), p. 9. 

51. Heinz Schiirmann, Der Paschamahlbericht Lk. 22, (7-14) 15-18 
(Munster, 1953), p. 35, Anm. 154. 

52. Joachim Jeremias, Die Abendsmahlsworte Jesu, 3te Aufl. (Gottingen, 
I960), pp. 174 f.; Eng. traas., The Eucharistic Words of Jesus (London, 1966), 
p. 183. 

53. Matthew Black, An Aramaic Approach to the Gospels and Acts, 
2nd ed. (Oxford, 1954), pp. 214 f.; 3rd ed. (1967), pp. 279 f. 

54. See, e.g., Ezra Abbot, "On the Reading 'Church of God', Acts xx. 
28," Bibliotheca Sacra, xxxiii (1876), pp. 313-52 (repr. in The Authorship of 
the Fourth Gospel and Other Critical Essays ^ [Boston, 1888], pp. 294-331); 

332 The Text of the New Testament 

elders: "Keep watch over yourselves and over all the flock, of which the 
Holy Spirit has made you overseers, to feed the church of . . . which 
he obtained with his own blood." Should the ellipsis be filled with the 
words "the Lord" or "God"? The external evidence is as follows: 

1. deov-. n, B, 218, 257, 383, 459, 614, 917, 1175, 1522, 1611, 1758, 
2138, 2298, al, ar, c, dem, ph, ro, w, Vulg, Syr p h , Athanasius, 
Basil, Ambrose, Epiphanius, Cyril of Alexandria. 

2. xvqiov. v 74 , A, C', D, E a , W, 33, 181, 209, 307, 337, 429, 431, 436, 
610, 623, 1739, 1891, e, gig, p, Sah, Boh, Syr* 1 mg , Arm, Iren lat , 
Didymus, Lucifer, Jerome, Pelagius. 

3. xvqiov xal deov: C 3 , H a , L a , P a , more than 100 minuscules, Old 
Slavonic, Theophylact. 

4. deov xou xvqiov. 47. 

5. xvqiov deov: 3, 95". 

6. Xqiotov: Syr eodd , Apost. Const, Athanasius codd . 

7. 'Irjoov Xqiotov. m. 

Of these seven readings, obviously only the first two merit any 
attention, for 6 and 7 are insufficiently supported and the rest are 
conflations of xvqiov and deov in various combinations preserved in 
Byzantine texts. 

On the basis of external evidence, it is difficult to decide which 
of the first two readings is original. Paleographically, the difference 
concerns only a single letter: KY and 0Y. Each is supported by early 
and diversified witnesses. Perhaps the most that can be said is that 
deov is the Alexandrian reading and xvqiov is supported by typical 
Western documents. One must rely chiefly on internal probabilities 
in reaching a decision. 

The phrase exxXrjoia xvqiov occurs seven times in the Septu- 
agint but nowhere else in the New Testament. On the other hand, 
exxlrjoia xov deov appears with moderate frequency (11 times) in 
the Epistles traditionally ascribed to Paul and nowhere else in the 
New Testament. (The expression ai exxXrjaiai Jtaoai xov Xqiotov 
occurs once in Rom. 16.16.) It is possible, therefore, that a scribe, 
finding deov in his exemplar, was influenced by Old Testament usage 
and altered it to xvqiov. It is also possible that a scribe, influenced by 
Pauline usage, changed the xvqiov of his exemplar to deov. 

Charles K De Vine, "The 'Blood of God' in Acts 20:28," Catholic Biblical 
Quarterly, ix (1947), pp. 381-408; as well as critical commentaries on the 
passage and the discussion by Hon in The New Testament in the Original 
Greek, ii, Appendix, pp. 98-100. 

The Practice of New Testament Textual Criticism 333 

One consideration that tips the scales in favor of the originality 
of the reading Oeov is the undeniable fact that deov is the more dif- 
ficult reading. The following clause refers to the church "that he 
obtained did. xov aifxaxoc, xov ldiox>." If this is taken in its usual sense 
("with his own blood"), the scribe might well raise the question 
"Does God have blood?" and thus be led to change Oeov to xvqlov. 
On the other hand, if xvqlov were the original reading, there is noth- 
ing unusual in the context that would cause a scribe to introduce the 
more difficult Oeov. This argument for deov as the older form is par- 
ticularly cogent if the change was made, as has recently been argued, 
in light of Patripassianist controversies that were raging in the early 
Church, when some heretical construals portrayed Christ as God 
himself (so that his death was the death of God). 55 

In this connection, it should be pointed out that did. xov atjuaxog 
xov idiov could have a meaning other than its usual one ("with his 
own blood"); this other meaning, which may have been in the mind 
of the writer of Acts, is "with the blood of his Own." (It is not nec- 
essary to suppose, with Hort, that viov may have dropped out after 
xov idiov, though paleographically such an omission would have 
been easy.) This absolute use of the singular number of idiog, which 
is otherwise unknown in the New Testament, is found occasionally 
in the Greek papyri as a term of endearment referring to near rela- 
tives.' 6 It is possible, therefore, that "his Own" (6 Xdiog) was a title 
that early Christians gave to Jesus, comparable with "the Beloved" (6 
dyannxog); compare Rom. 8.32, where Paul refers to God "who did 
not spare xov idiov viov" in a context that clearly alludes to Gen. 
22.16, where the Septuagint has xov aycutrjxov viov. 

In conclusion, whatever may be thought of the slight possibility that 
xov idiov is used here as the equivalent of xov idiov viov, it appears that 
the reading Oeov was more likely to have been altered to xvqiov than 
vice versa, and it is therefore to be regarded as the original. 57 

55. See Bart D. Ehrman, The Orthodox Corruption of Scripture: The 
Effect of Christological Controversies on the Text of the New Testament (New 
York, 1993), pp. 87-8. 

56. James Hope Moulton, A Grammar of New Testament Greek, i, Pro- 
legomena, 3rd ed. (Edinburgh, 1908), p. 90. 

57. The reading deov is preferred by Westcott and Hort, Weiss, Vogels, 
Merk, Bover, von I larnack (Beitrdge, iv, p. 330), Jackson and Lake (Begin- 
nings of Christianity), G. H. C. Macgregor (Interpreter's Bible), F. F. Bruce 
(Acts), and C. S. C. Williams (Harper's [= Black's] New Testament Commen- 
taries), De Vine, and Ehrman. 

334 The Text of the New Testament 

Of all the letters attributed to Paul, the letter to the Colossians 
contains proportionately the greatest number of textual problems. 
The close of Col. 2.2 presents what is, at first, a bewildering variety 
of readings; the manuscripts present 15 different conclusions of the 
phrase eig imyvcooiv xov juvoxtjqlov ("to the knowledge of the 
mystery of): 

A 1. xov OeovXqloxov: p 46 , B, Vulg ms , Hilary of Poitiers, Pelagius. 

2. xov Oeov-. D pl , HP, PP, 69, 424 c , 436, 462, 1912, Sah MS . 

3. xov Xqloxov-. 81, 1241, 1462, (1739), b. 

4. xov Oeov o eoxiv Xoioxog: D p *, ar, d, o, Vulg mss , Augustine. 

5. xov Oeov o eoxiv tteqi Xqloxov-. Eth. 

6. xov Oeov xovev Xqloxw: 33, (Clement), Ambrst. 

7. xov Oeov xovev Xqioxw 'Irjoov-. Arm. 

8. xov Oeov xai Xqioxov-. Cyril of Alexandria. 
B 9- tov Oeov naxgog Xqioxov: «*, 048, 216, 440. 

10. xov Oeov jcaxQog xov Xqioxov-. A, C, 0150, 4, 1175, Sah, Boh. 

11. xov Oeov xai Jiaxgog xov Xqloxov-. s c , L, W, 256, 263, pc, 
Vulg ms , Slav. 

12. xov 6eov naxgog xai xov Xqioxov-. 075, 0208, 441, 1908, 
SyrP, Chr. 

13- tov Oeov naxQog xai xov Xqioxov lr\oov: Vulg, Boh codd , 

14. xov Oeov naxQog xai xov xvqiov fjfj3v Xqloxov 'Itjoov-. 1, 
Vulg MS . 

15. tov Oeov xai naxgbg xai xov Xqloxov-. D p(c) , E p , K p , IP, most 
minuscules, Syr 11 , Theod. 

Of all these variant readings, the one that has been placed first is 
to be preferred on the basis of both external and internal considera- 
tions. Externally, it is supported by the earliest and best Greek man- 
uscripts; internally, the difficulty of the expression xov fivoxrjQiov 
xov Oeov Xqloxov has led to a multitude of scribal interpretations. 
An obviously popular expedient was insertion of the word JtaxQog; 
this addition appears in seven of the variant readings (those grouped 
under B). Insertion of the article before Xqioxov (readings 10-15) is 
plainly in the interest of making the expression parallel with xov 
Oeov. The reading placed last in the list (it lies behind the rendering 
of the King James Version, "the mystery of God, and of the Father, 
and of Christ"), though supported by the largest number of wit- 
nesses, is also the weakest, for it is a conflation of the two types of 
amelioration represented in readings 11 and 12. 

The Practice of New Testament Textual Criticism 335 

If reading 9 were original, then the rise of all eight readings 
grouped under A is inexplicable, for why should Jiargog have fallen 
out? On the contrary, naxgoc, was inserted in order to clarify the syn- 
tactical relation between deov and Xqiotov (for reading 1 could 
mean "the knowledge of the mystery of God Christ," "the knowledge 
of the mystery of God's Christ," or "the knowledge of God's mystery, 
of Christ"). Besides the insertion of jzargog (readings 9-15), several 
other attempts were made to explain the relationship of Xqiotov to 
Oeov (readings 4-8). The scribes responsible for readings 2 and 3 
sought to relieve the difficulty by the elimination of one or the other 
of the two genitives, and in support of 3 the scribe could point to 
Eph. 3-4 as a precedent (rq> fivOTrjQicp tov Xqiotov). Reading 4 gives 
what must be the right sense, suggesting that in reading 1 the word 
Xqiotov is explanatory of tov fivoxrjQiov xov deov. Perhaps in 
dictating the epistle the author separated the word Xqiotov from the 
preceding phrase by a slight pause for breath, which can be repre- 
sented in modern printing by a comma. Thus, it is possible to 
explain the origin of all the other readings on the assumption that 
reading 1 is original, but this reading cannot be described as deriva- 
tive of any of them. Since the external support of reading 1 is the 
best, regarding both age and character, one must conclude that tov 
deov Xqiotov is the earliest attainable form of text preserved among 
the extant witnesses. 58 

As was remarked earlier, the harmonization of the Evangelists, 
whether done by scribes deliberately or unconsciously, is by defini- 
tion a secondary process. Therefore, the supreme rule for editors of 
the text is to give each Gospel its own proper character. Of many, 
many examples that might be cited as illustrations of this basic prin- 
ciple, the following has been chosen to show that on very rare 
occasions the correct reading may be preserved alone in the Koine 
or Byzantine text. 

If one were to consider only the external evidence in support of 
aQxi£Q£iQ in Luke 20.1 ("the chief priests and the scribes with the 
elders came up"), there would appear to be no doubt whatever con- 
cerning its originality: the diversity and the antiquity of the witnesses 

58. The caution with which the conclusion is expressed is in defer- 
ence to those who, like F. W. Beare in the Interpreter's Bible, think that the 
expression xov deov Xqiotov is too difficult for the author to have written 
and that some primitive error must lie behind the earliest attainable text. 

336 The Text of the New Testament 

are most impressive, and the support for the variant reading legelg 
appears to be negligible. Thus: 

1. dgxiegelg: s, B, C, D, L, N, Q, 0, W, fam. 1, (fam. 13), fam. 1424, 
33, 157, 579, 892, 1071, 1241, 1604, Old Lat, Vulg, Syr, Sah, Boh, 
Arm, Geo, Eth, Diatess Arab ' Ilal(VenX Dutch < L *i5<--> 

2. legelg: A, E, G, H, K, S, U, V, W, r, A, A, IJ, many minuscules, 

A Gospel harmony, however, shows that the parallels to this 
account (Matt. 21.23 and Mark 11.27) refer to "the chief priests" in 
company with scribes and/or elders, with no fluctuation in the man- 
uscript evidence. Furthermore, a concordance provides the informa- 
tion that the same kind of stereotyped formula appears nearly 
50 times throughout the Gospels and Acts, and in no case has legelg 
replaced dgxiegelg. 

In the passage under consideration, there is no discernible mo- 
tive for altering "chief priests" to "priests," whereas the influence of 
the stereotyped phrase containing dgxiegelg would have been felt 
independently by scribes of manuscripts in every textual tradition. 
One must conclude, therefore, that Tischendorf and von Soden are 
correct in printing legelg in Luke 20.1 and that Westcott and Hort, 
Weiss, Vogels, Merk, Bover, Nestle, the British and Foreign Bible 
Society's text of 1958, and the United Bible Societies' fourth edition 
are wrong in preferring dgxiegelg . 59 

There are other instances besides Luke 20.1 where almost all of 
the "good" manuscripts are in error and the correct reading is pre- 
served in "inferior" witnesses. Thus, at Matt. 12.47, the best Alexan- 
drian witnesses (B, K*), joining with representatives of the Western 
group iff 1 , k, Syr sc ), are convicted of having accidentally omitted a 
whole verse. Verse 47 closes with the same word as verse 46, and the 
eyes of copyists have chanced to wander from the end of one to the 
end of the other. 

In Heb. 7.1, the relative pronoun oq which is not suited to the 
syntax of the sentence, is supported by », A, B, C c , D p , I, K p , 33, and 
al. The grammatically correct 6 is witnessed by p 46 , C*, V, 1739, and 
the Koine group of manuscripts. In this case, one can see how the 

59. For another discussion of this passage, see Heinrich Greeven's 
comments in New Testament Studies, vi (I960), pp. 295 f., where he comes 
to the same conclusion as that adopted here. 

The Practice of New Testament Textual Criticism 337 

primitive error entered both the Alexandrian and Western traditions: 
the following word begins with the letter sigma (ovvavTijoag), so 
the correct 6 came to be pronounced and spelled og. 

Occasionally, considerations relating to intrinsic evidence will 
cast a decisive vote in the face of what appears to be overwhelming 
external testimony. The generally received text of Matt. 22.34-5 
reads "When the Pharisees heard that he (Jesus] had silenced the Sad- 
ducees, they gathered together, and one of them, a lawyer, asked 
him a question to test him." This all seems straightforward enough 
and quite unexceptional. There are, however, a few witnesses that 
omit vo/iixog ("a lawyer") at verse 35, namely, fam. 1, e, Syr s , Arm, 
Geo, and Origen according to Rufinus' Latin translation. Normally, 
this evidence, comprising only part of the Caesarean text with some 
versional support, would not be regarded as especially weighty. In 
this case, however, internal considerations of the most compelling 
kind make it altogether likely that all of the majuscules and almost 
all of the minuscules are wrong in reading vofiixog. In the first place, 
except for this passage, it is Luke alone of the four Evangelists who 
employs the word vofuxoq; the usual word in Matthew for one 
concerned with the Mosaic law is yga/ufiarevg, which appears 
frequently in all three Synoptic Gospels. In the second place, an ex- 
amination of a harmony of the Gospels discloses that in the parallel 
passage Luke reads "Just then a lawyer stood up to test Jesus" 
(10.25). There is no reason why scribes should have omitted vofiixog 
from Matt. 22.35, but there would have been a strong temptation to 
add such a circumstantial detail from the parallel account in Luke. 
Despite, therefore, the rather limited range of evidence in support of 
the shorter text, it appears that vofiixog is an intrusion into the text 
of Matt. 22.35. 

In 1 Thess. 3-2, "We sent Timothy, our brother and ... in the 
gospel of Christ, to establish you in your faith and to exhort you," the 
word or words after "our brother and" have been transmitted in five 
different forms: 

1. "God's fellow worker" (ovvegyov xov deov: D p *, 33, b, d, mon, o, 
Pelagius, Ambr, pseudo-Hieronymus). 

2. "fellow worker" (ovvegyov-. B, Eph vid ). 60 

60. See Joseph Molitor, Der Paulustext des hi. Ephrdm aus seinem 
armenisch erhaltenen Paulinenkommentar (Rome, 1938), p. 112. 

338 The Tf.xt of the New Testament 

3. "God's servant" (didxovov xov deov-. «, A, P p , V, 424 c , 1739, Goth, 
Cop, Arm, Basil, Theod, Mops lat ). 

4. "servant and God's fellow worker" {diaxovov xal ovvegyov xov 
Oeov: FP, GP). 

5. "God's servant and our fellow worker" (diaxovov xov Oeov xal 
ovvegyov rjficbv. D 1 *^, K p , IP, Syr p h , Eth, most minuscules, Chr, 

The reading placed last in the list, though supported by the great 
majority of witnesses, is obviously a secondary one that was formed 
by joining readings 3 and 2 with the addition of the word rj/udiv. This 
conflate reading, so typical of the Koine type of text, lies behind the 
rendering of the verse in the King James Version. 

The fourth is also a type of conflated reading. The reviser has 
rather mechanically combined readings 3 and 1, writing the qualify- 
ing genitive xov Oeov only once. The external support for this vari- 
ant is of little value, for the two bilingual manuscripts F and G are so 
closely related to each other as to constitute but one witness, which 
belongs to a secondary stratum of the Western type of text. We are 
left therefore with readings 1-3. Which of these gave rise to the 
others? It is difficult to answer this question with assurance, for, as it 
happens, more or less compelling reasons can be found for prefer- 
ring each of the three readings as the original, reasons that led Bover 
and the translators of the New English Bible to adopt reading 1, 
Weiss and the New Revised Standard Version to adopt 2, and 
Tischendorf, Westcott and Hort, von Soden, Vogels, Merk, and the 
Revised Standard Version to adopt 3- Arguments that support each of 
these choices involve the following considerations. 

The second reading, ovvegyov, can scarcely be original, for then 
one would have no adequate explanation for the addition of xov 
Oeov, producing reading 1, or its alteration to diaxovov with the 
addition of xov Oeov, producing reading 3- 

It cannot be disputed that the third reading, "God's servant," is 
attested by several excellent Alexandrian witnesses as well as by rep- 
resentatives of widely diversified versions. But the expression is both 
so clear and so rich that, if original, no one would have felt the need 
of altering it. The same cannot be said of the first reading in the list, 
for the application to Timothy of the expression "God's fellow 
worker" may have offended the scrupulous. How can any person 
(so some may have reasoned) be given so exalted a title, least of all 
a young convert who was merely assisting Paul? Two courses were 
open to those who wished to soften the idea: by suppressing 

The Practice of New Testament Textual Criticism 339 

xov deov, Timothy becomes Paul's fellow worker (reading 2), in 
accordance with his designation in Rom. 16.21, or by keeping xov 
deov and substituting biaxovov for ovvegyov, Timothy is again 
described (reading 3) in typically Pauline language (Rom. 13-4, 
2 Cor. 6.4). Thus, it would appear that reading 1, ovvegyov xov 
deov, is original, for it explains the origin of the others. 

One's confidence in the preceding arguments, however, may 
well be weakened by a second review of the evidence. Those who 
are moved chiefly by considerations based on external evidence will 
certainly not be impressed by the combination of D p *, 33, b, mon, o, 
Pelagius, Ambrosiaster, and pseudo-Hieronymus. Furthermore, the 
argument that was used above to discredit the originality of reading 
ovvegyov xov deov, namely, that it would have offended the 
scrupulous, becomes less cogent when one observes that a similar 
expression in 1 Cor. 3-9 (deov ydg iofiev ovvegyoi) has occasioned 
no textual variation in any known witness. Furthermore, the gener- 
ally acknowledged tendency of Western texts to substitute syn- 
onyms and near-synonyms suggests that reading 1 arose as a scribal 
alteration of 3- 

The chief weakness, however, of the argument in favor of the 
originality of reading 3 is the absence of any good reason to account 
for the deletion of xov deov to produce the second reading. Weiss 
therefore argues that the simplest explanation to account for the rise 
of the many forms of the text is to regard ovvegyov of 13 as the 
original, of which all the others are modifications. 61 

In assessing the argument supporting each of the variant read- 
ings, one will do well to exercise great caution. According to 
Lachmann, the degree of probability attached to any particular vari- 
ant may be assessed as certain, almost certain, probable, doubtful, 
very doubtful, or nil. In this case, whatever choice is made among 
the three chief variants, every candid critic must acknowledge the 
strength of the arguments supporting the other two. 

Another example where conflict between external and internal 
evidence makes it impossible to come to a confident decision among 
variant readings is found in Acts 12.25- The natural impression that 
one gets when reading the section 11.27 to 13.1 is that 11.30 refers 
to the arrival of Paul and Barnabas at Jerusalem and that 12.25 ought 

61. B. Weiss, Textkritik der paulinischen BriefeiTexte und Unter- 
suchungen, xiv (3); Leipzig, 1896), pp. 13 f- It is significant that in this 
case Westcott and Hort did not see fit to follow their favorite manuscript. 

340 The Text of the New Testament 

to tell of their departure from Jerusalem. There is, however, early and 
rather widespread support for the almost impossible reading eig 
'iEQOvoakfini*, B, H a , L a , P a , 81, Syr 11 m s, Sah ms ). 62 The other readings 
are i^lEQOvoaXj]/i (p 74 , A, 33, 547, 913, 1739, 1891, Sah mss , Arm, Eth) 
and and 'IegovoaX^fi (D, [E], W, 181, 431, 614, ar, c, d, dem, gig, ph, 
ro, Vulg; Syr p witnesses to either £§ or and). 

All the canons of textual criticism favor the more difficult read- 
ing, eig, supported as it is by the earliest and best manuscripts; but 
the sense of the context cries out for a preposition that means 
"from." 63 Was eig a primitive corruption that later scribes attempted 
to correct by altering it to e£ or curd? Here, we must acknowledge 
that we simply do not know what the author originally wrote. 64 

Does Luke say that Jesus sent out 70 or 72 disciples (Luke 10.1, 
17)? The evidence is as follows. 

ipdofirjxovra is supported by p 45 (extant for verse 17 only) , 65 s, 
A, C, L, W, A, 0, A, E, W, fam. 1, fam. 13, most minuscules,/ q, i 
(extant for verse 1 only), Syr c (vs m ?• h ' P al , Cop boh , Goth, Iren, Tert, 
Clem, Orig, Eus, Ambr, Jerome. 

62. This Sahidic manuscript is B.M. Or. 7594, thought by some to date 
from the fourth century; see n. 94 on p. 110. 

63. According to the addenda to Tischendorfs 8th ed. (as reported in 
Gregory's Prolegomena, p. 1281), the scribe of B had begun to write and 
but then changed it to etc. 

64. Hon (Westcott and Hon, The New Testament in the Original 
Greek, p. 94) suggested that the passage be emended by transposing the 
order of the words, thus separating etc 'Iegovoalijft from vneargexpav and 
ranging it with what follows (vjieorgerpav zfjv eig legovaakrjfi 
TiXnowoavxeg diaxoviav, "they returned, having fulfilled their ministry at 
Jerusalem"). Dom J. Dupont concluded a lengthy examination of the verse 
by suggesting that the preposition eig be retained and the difficulty in 
sense be alleviated by punctuating with a comma after the main verb 
(vjteoTQeipav, eig 'IegovoaXrifi Ttkngwoavxeg rfjv diaxoviav); see his arti- 
cle, "La Mission de Paul 'a Jerusalem' (Actes xii. 25)," Novum Testamentum, 
i (1956), pp. 275-303. 

65. F. G. Kenyon erroneously reported the evidence of p 45 as support- 
ing the numeral 72 (off). Bruce Metzger made an examination of this 
passage in p 45 under natural and artificial light and has assured himself that 
the Greek character that follows the omicron is nothing but a diple, or 
space-filler (>), which scribes used occasionally in order to bring an 
otherwise short line even with the right-hand margin of the column. 

The Practice of New Testament Textual Criticism 341 

efidofiijxovTa dvo is supported by p 75 , B, D, M (verse 1 only), 
0181 (verse 1 only), a, b, c, d, I, r 2 , Vulg, Syr s - c(vs - 1) - hni « (V5 - 17) , Cop sah , 
Arm, Georg 12 , Diatess Du,ch Pers Ilal , Orig, Ephr, Aug, Ambrst. 

The external evidence is almost evenly divided: the chief repre- 
sentatives of the Alexandrian and the Western groups, with most of 
the Old Latin and the Sinaitic Syriac, support the numeral 72; how- 
ever, other Alexandrian evidence of relatively great weight (k, L, A, 
A, S) as well as Caesarean witnesses (p 45 , fam. 1, fam. 13) join in 
support of the numeral 70. 

The factors bearing upon the evaluation of internal evidence, 
whether involving transcriptional or intrinsic probabilities, are singu- 
larly elusive. It is likely that in most of the early manuscripts (as in p 45 
and p 75 ) the numeral was written with letters of the alphabet (either 
o/3 or o). It was easy, therefore, for either number to be accidentally 
altered to the other. If the variation was made deliberately, one can 
imagine that an Alexandrian scribe with a mathematical penchant 
altered 70 to 72 for the sake of what he may have regarded as schol- 
arly symmetry. On the other hand, if the alteration was made unwit- 
tingly, it is perhaps more likely that the precise number should 
be transformed into the round number 70 than that the "solemn" 
number 70 should be transformed into 72. 

Those who transmitted the account prior to its inclusion in Luke 
10 may have wished to convey a symbolic meaning in the number of 
the disciples, and it is easy to find parallels in Jewish antiquity for ei- 
ther 70 or 72. Seventy elders were appointed by Moses to assist him 
(Num. 11.16-17, 24-5); there were 70 sons of Jerubbaal (Judges 9.2), 
70 sons of Ahab (2 Kings 10.1), and 70 priests of Bel (Bel and 
Dragon, verse 10). 

On the other hand, according to the Letter of Aristeas (§§ 46-50), 
72 elders (six from each of the 12 tribes) were chosen in order to 

In fact, by consulting Kenyon's volume of plates of p 4 \ anyone can see the 
similarity between the disputed character and the cliple that appears on the 
same folio near the top of the column. For a fuller discussion of the reading 
of p 4 ', including C. H. Roberts' view that the numeral is 76 (og), see B. M. 
Metzger, "Seventy or Seventy-two Disciples?", New Testament Studies, v 
(1959), pp. 299-306, on the basis of which the citation of the evidence of 
P'" was corrected in the I960 edition of Nestle's Greek Testament. Other 
editions of the Greek Testament continue to repeat the error in Kenyon's 
edition of p 45 . 

342 The Text of the New Testament 

prepare a Greek translation of the Torah (the Septuagint), and in 3 
Enoch the number of princes of kingdoms on high is 72, corre- 
sponding to the 72 languages of the world (17.8; cf. 18.2 f., 30.2). 66 

It is, however, exceedingly difficult to ascertain what symbolism 
is intended in Luke's account. On the one hand, if the mission of this 
group of disciples is to be understood as a mission to Israel, the 
number may have been chosen as a multiple of the 12 tribes of 
Israel. On the other hand, since several New Testament writers pre- 
suppose a parallel between Jesus and Moses, 67 it may be that this 
group of Jesus' disciples is intended to correspond to the elders who 
assisted Moses. So evenly balanced are these two possibilities that it 
is hazardous to dogmatize as to which is more probable. 

A total appraisal of both the external and internal evidence 
bearing on these variant readings must remain indecisive. Though 
the reading "seventy-two" is supported by a combination of early 
witnesses that normally carries a high degree of conviction of origi- 
nality, yet the witnesses that read "seventy" are so weighty and the 
internal considerations so evenly balanced that the textual critic must 
simply acknowledge an inability to decide with assurance between 
the two. 

The passages discussed above will suffice to illustrate the wide 
variety of considerations that one must take into account when eval- 
uating variant readings. In addition to solving such problems indi- 
vidually as they occur in the text, an editor of the Greek Testament 
is confronted with still others arising from the diversity of usage and 
authorship within the New Testament as a whole. How far, for 
example, should an attempt be made to standardize first aorist end- 
ings on second aorist stems? Shall an edition reproduce divergences 
between two or more authors in the spelling of the same proper 
name? Or, even within the same book, is it permissible to vary the 
spelling of a proper name? For example, should one print the title of 
Paul's Epistle to the Colossians FIFOS KOAA22AEI2 (with p 46 , A, 
B*, I, K p , and many manuscripts), as the Textus Receptus, Lachmann, 
and Westcott and Hort do, but in Col. 1.2 spell the name of the city 

66. For many other examples from Jewish antiquity involving 70 and 
72, see Metzger, "Seventy or Seventy-two Disciples?" 

67. See J. Jeremias in Kittel, Theological Dictionary of the New 
Testament, iv, pp. 867-73. 

The Practice of New Testament Textual Criticism 343 

Kokoooalg (with p 46 , H, [A has a lacuna], B, D p , EP, FP, G p , LP, etc.)? 
It is far easier to ask such questions than to answer them. 

By way of conclusion, let it be emphasized again that there is no 
single manuscript and no one group of manuscripts that the textual 
critic may follow mechanically. All known witnesses of the New 
Testament are to a greater or lesser extent mixed texts, and even sev- 
eral of the earliest manuscripts are not free from egregious errors. 
Although in very many cases the textual critic is able to ascertain 
without residual doubt which reading must have stood in the 
original, there are not a few other cases where only a tentative deci- 
sion can be reached, based on an equivocal balancing of probabilities. 
Occasionally, none of the variant readings will commend itself as 
original, and one will be compelled either to choose the reading that 
is judged to be the least unsatisfactory or to indulge in conjectural 
emendation. In textual criticism, as in other areas of historical re- 
search, one must seek not only to learn what can be known but also 
to become aware of what, because of conflicting witnesses, cannot 
be known. 


The following is a chronological list of introductions to New Testa- 
ment textual criticism. In addition to these volumes, many dictionar- 
ies of the Bible, introductions to the New Testament, and histories of 
the English Bible contain brief surveys of the subject. For a bibliogra- 
phy on special points of New Testament textual criticism, reference 
may be made to monographs mentioned in the appropriate sections 
above as well as to two other publications by Bruce M. Metzger: 
Annotated Bibliography of the Textual Criticism of the New Testament, 
1914-1939 (Copenhagen, 1955) and Chapters in the History of New 
Testament Textual Criticism (Grand Rapids, MI, 1963). 

J. I. Doedes, Verhandeling over de tekstkritiek des Nieuwen Verbonds (Teyler's 

godgeleerd Genootschap, deel xxxiv), Haarlem, 1844, 481 pp. 
J. Scott Porter, Principles of Textual Criticism with their Application to Old 

and New Testaments, London, 1848, 515 pp. 
Thomas Sheldon Green, A Course of Developed Criticism on Passages of the 

New Testament Materially Affected by Various Readings, London, 1856, 

192 pp. 
Samuel Prideaux Tregelles, Introduction to the Textual Criticism of the New 

Testament; being vol. iv of Thomas Hartwell Home's An Introduction to 

the Critical Study and Knowledge of the Holy Scriptures, 10th ed., London, 

1856; 13th ed., London, 1872, 402 pp. 
F. H. A. Scrivener, A Plain Introduction to the Criticism of the New Testament, 

London, 1861, 490 pp.; 2nd ed., Cambridge, 1874, 607 pp.; 3rd ed., 

Cambridge, 1883, 712 pp.; 4th ed., 2 vols., ed. by Edward Miller, 

London, 1894, vol. 1, 418 pp., vol. 2, 428 pp. 
C. E. Hammond, Outlines of Textual Criticism Applied to the New Testament, 

Oxford, 1872, 138 pp.; 6th ed., revised, Oxford, 1902, 179 pp. 
William Milligan and Alexander Roberts, The Words of the New Testament, 

as Altered by Transmission and Ascertained by Modern Criticism, 

Edinburgh, 1873, 262 pp. 



Frederic Gardiner, Principles of Textual Criticism; with a Graphic Table of 

Uncials, reprinted from Bibliotheca Sacra, xxxii (1875), pp. 209-65. 
F. H. Scrivener, Six Lectures on the Text of the New Testament and the Ancient 

Manuscripts Which Contain It, Chiefly Addressed to Those Who do not Read 

Greek, London, 1875, 216 pp. 
Thomas Rawson Birks, Essay on the Right Estimation of Manuscript Evidence 

in the Text of the New Testament, London, 1878, 128 pp. 
Brooke Foss Westcott and Fenton John Anthony Hort, Introduction, 

Appendix, in The New Testament in the Original Greek, ii, London, 1881, 

324, 188 pp.; 2nd ed., 1896, 330, 180 pp. 
J. P. P. Martin, Introduction a la critique textuelle du Nouveau Testament: Partie 

theorique, Paris, c. 1883, 712 pp.; Partie pratique, 5 vols., Paris, c. 1883-6, 

pp. 327, 554, 512, 549, 248. 
Philip Schaff, A Companion to the Greek Testament and the English Version, 

New York, 1883, 616 pp.; 3rd ed., revised, New York, 1889, 618 pp. 
Fr. Schjott, "Det ny Testamentes Texthistorie i de tre forste Aarhun- 

dreder," Teologisk Tidsskriftfor den danske Folkekirke, i (1884), pp. 343-92; 

"De nytestamentlige Uncialhaandskrifter," Teologisk Tidsskrift for den 

danske Folkekirke, vi (1889), pp. 432-57, 500-55. 
Caspar Rene Gregory, Prolegomena, being vol. Hi of Tischendorf's 

Novum Testamentum Graece, ed. octava critica maior, Leipzig, 1884-94, 

1426 pp. 
Edward Miller, A Guide to the Textual Criticism of the New Testament, 

London, 1886, 147 pp., reprinted 1979. 
Benjamin B. Warfield, An Introduction to the Textual Criticism of the New 

Testament, London, 1886, 225 pp. 
Arnold RueGG, Die neutestamentliche Textkritik seit Lachmann; ein Versuch 

zur Orientierung, Zurich, 1892, 97 pp. (with a list of N.T. MSS., their 

editors, and collators). 
Eberhard Nestle, Einfuhrung in das griechische Neue Testament, Gottingen, 

1897, 129 pp.; 2te Aufl., Gottingen, 1899, 288 pp.; 3te AufL, Gottingen, 

1909, 298 pp.; Eng. trans, from 2nd German ed. by William Eadie, 

Introduction to the Textual Criticism of the Greek New Testament, London, 

1901, 351 pp. 
Charles F. Sitterly, Praxis in Manuscripts of the Greek Testament, New York, 

1898; 4th ed. is part 3 of The Canon, Text and Manuscripts of the New 

Testament, New York, 1914, 126 pp. 
Matheus Lundborg, Nya Testamentets text, dess historia och kritiska behan- 

dling i allmanna grunddrag, Lund, 1899, 406 pp. 
Marvin R. Vincent, A History of the Textual Criticism of the New Testament, 

New York, 1899, 185 pp. 
Kirsopp Lake, The Text of the New Testament, Oxford, 1900, 104 pp.; 6th ed., 

revised by Silva New, Oxford, 1928, 104 pp. 
Frederic G. Kenyon, Handbook to the Textual Criticism of the New Testament, 

London, 1901, 321 pp.; 2nd ed., London, 1912, 381 pp. 

Bibliography 347 

Rudolf Knopf, Der Text des Neuen Testaments, Giessen, 1906, 48 pp. 
August Pott, Der Text des Neuen Testaments nach seiner geschichtlichen 

Entwicklung, Leipzig, 1906, 108 pp.; 2te Aufl., Leipzig, 1919, 116 pp. 
Caspar Rene Gregory, Canon and Text of the New Testament, New York, 1907, 

pp. 297-539. 

, Textkritik des Neuen Testamentes, 3 vols., Leipzig, 1900-9, 1486 pp. 

Hermann Freiherr von Soden, Die Schriften des Neuen Testaments in ihrer 

altesten erreichbaren Textgestalt, I. Teil, Untersuchungen, i. Abteilung, 

Die Textzeugen, Berlin, 1902, 704 pp.; ii. Abteilung, Die Textformen, a. 

Die Evangelien, 1907, pp. 705-1520; iii. Abteilung, Die Textformen, b. Der 

Apostolos mit Apokalypse, 1910, pp. 1521-2203. 
Jens Stub Irgens, De trykte grseske Nye Testamenters historie tilligemed en 

indledning dertil og et anhang, Kristiania, 1907, 196 pp. 
Ernest Jacquier, he Nouveau Testament dans I'eglise chretienne. II. he Texte du 

Nouveau Testament, Paris, 1913, 535 pp. 
Alexander Souter, The Text and Canon of the New Testament, London, 1913, 

pp. 1-145; revised by C. S. C. Williams, London, 1954, pp. 1-133. 
P. G. Groenen, Algemeene inleiding tot de Heilige Schrift. II. Geschiedenis van 

den tekst, Leiden, 1917, 375 pp. 
Ernst von Dobschutz, Eberhard Nestle's Einfiihrung in das griechische Neue 

Testament, 4te Aufl. vollig umgearbeitet, Gottingen, 1923, 160 pp. 
Heinrich Joseph Vogels, Handbuch der Textkritik des Neuen Testaments, 

Munster, 1923, 255 pp.; 2et Aufl., Bonn, 1955, 236 pp. 
A. T. Robertson, An Introduction to the Textual Criticism of the New Testament, 

New York, 1925, 300 pp.; 2nd ed., New York, 1928, 300 pp. 
Gfuseppe Sacco, La Koine del Nuovo Testamento e la trasmissione del sacro testo, 

Rome, 1928, pp. 151-327. 
Ernst Nachmanson, Nya Testamentet, en oversikt av dess yttre historia, 

Stockholm, 1931, 164 pp. 
Leon Vaganay, Initiation a la critique textuelle neotestamentaire, Paris, 1934, 

188 pp.; English trans., London, 1937, 208 pp; 2nd ed. by Christian- 
Bernard Amphoux, Paris, 1986, 300 pp.; English trans, by Jenny 

Heimerdinger, Cambridge, 1991, xxiv + 227 pp. 
M.-J. Lagrange, Introduction a I'etude du Nouveau Testament; 2nd part, 

Critique textuelle. II. La critique rationnelle, Paris, 1935, 685 pp. 
Eugene Mercier, Le Texte du Nouveau Testament, Lausanne, 1935, 127 pp. 
Auguste Hollard, Histoire du texte du Nouveau Testament, Paris, 1936, 80 pp. 
Frederic G. Kenyon, The Text of the Greek Bible: A Students' Handbook, 

London, 1937, 264 pp.; 2nd ed., London, 1949, 264 pp.; German trans. 

by Hans Bolewski, Der Text der griechischen Bibel: ein Lehrbuch, 

Gottingen, 1952, 166 pp.; 2te Aufl., uberarbeitet und erganzt von 

A. W. Adams, 1961, 200 pp.; 3rd English ed., revised by A. W. Adams, 

London, 1975, 275 pp. 
R. Wehner, Nya Testamentets grundtext genom seklerna, Stockholm, 1943, 

34 pp. 

348 Bibliography 

Johannes Sundwall, Nya Testaments urtext, Abo, 1946, 104 pp. 

Ernest Cadman Colwell, What Is the Best New Testament? Chicago, 1952, 

127 pp. 
Paolo Sacchi, Alle origini del Nuovo Testamento: Saggio per la storia della 

tradizione e la critica del testo, Florence, 1956, 178 pp. 
L. D. Twilley, The Origin and Transmission of the New Testament: A Short 

Introduction, Edinburgh, 1957, pp. 36-63. 
Jean Duplacy, Ou en est la critique textuelle du Nouveau Testament? Paris, 

1959, 103 pp. 
A. Ivanov, "Tekstual'nye pamyatniki svyashchennykh novozavetnykh 

pisanil," Bogoslovskie trudy, i (1959 [I960]), pp. 53-83. 
Vincent Taylor, The Text of the New Testament: A Short Introduction, New 

York, 1961, 113 pp. 
TOSHIO Hirunuma, New Testament Textual Criticism [in Japanese], Tokyo, 

1962, 192 pp. 
J. Harold Greenlee, An Introduction to New Testament Textual Criticism, 

Grand Rapids, MI, 1964, 160 pp.; London, 1976; 2nd ed., Peabody, MA, 

Bruce M. Metzger, The Text of the New Testament: Its Transmission, Corrup- 
tion, and Restoration, Oxford, 1964, xii + 270 pp.; German trans., 

Stuttgart, 1966; Japanese trans., Tokyo, 1973, 2nd ed., 1999; Chinese 

trans., Taipei, 1981; Korean trans., Seoul, 1984; Italian trans., Brescia, 

1996; Russian trans., Moscow, 1996. 
Harold K. Moulton, Papyrus, Parchment and Print; the Story of how the New 

Testament Text has Reached Us (World Christian Books 57), London, 

1967,77 pp. 
Jack Finegan, Encountering New Testament Manuscripts: A Working Introduc- 
tion to Textual Criticism, Grand Rapids, MI, 1974, 203 pp. 
J. J. Thierry, Korte geschiedenis van tekst van het Nieuwe Testament, Kampen, 

1982, 132 pp. 
Kurt Aland and Barbara Aland, Der Text des Neuen Testament, Stuttgart, 

Grand Rapids, MI, 1982; English trans, by E. F. Rhodes, Grand Rapids, 

MI, 1987, 338 pp.: 2nd ed., Grand Rapids, MI, 1989, 366 pp. 
Kobus (i.e., J. H.) Petzer, Die teks van die Nuwe Testament (Hervormde Teolo- 

giese Studies, Supplementum 2), Pretoria, 1990, xviii + 353 pp. 
Wilson Paroschi, Critica Textual do Novo Testamento, Sao Paulo, 1993, 248 pp. 
David Alan Black, New Testament Textual Criticism, a Concise Guide, Grand 

Rapids, MI, 1994, 79 pp. 
Keith Elliott and Ian Moir, Manuscripts and the Text of the New Testament, 

An Introduction for English Readers, Edinburgh, 1995, x + 111 pp. 

General Index 

Page numbers set in italics refer to figures. 

Abad, j. M., 139n4 

Abbot, ¥... 150n32, 331n54 

Abbott, T. K., 81.87, 88 

Abbreviations, 17. 46 

Abraham. Bishop, 127nl23 

Abuladze, I., 119 

Accidental scribal omissions, theory of, 

Account of the Printed Text of the Creek New 

Testament, An (Tregelles). 174 
Achtemeier, P. J., 23n28 
Adler.J. G. C, 168 
Adoptionist Christological views, 285 
Adysh manuscript, 118-19 
African manuscript groups, 159, 161 
African Old Latin manuscripts, 101, 102 
Agrapha, 268 
Aitken, R., 163n64 
Ajamian, S.. 118 
Aland, B.. 18n23, 99, 191, 194, 237-38, 

Aland, K., 47n75, 59nl4, 64nl8, 105. 110n93, 

115nl03, 117nl05, 121nlll, 143nl6, 

147, 185n29. 190, 191, 192n46, 193, 

221n34, 233n62, 237-38, 277n8, 308nl6, 

Albors, P. B„ 127nl23 
Akuin, 106 
Aldine Press, 149 
Aldred (priest), 125 
Alexander (monk), 201 
Alexander, C, 164 
Alexander, J. J, G., 43n62 
Alexander II, Czar of Russia, 63-64 
Alexandrian recension, 161-62, 165-66 
Alexandrian text, 52, 169. 170, 187, 190, 221, 

307, 315, 318, 327, 329. 337, 338, 341 
rise and development of, 277-78 
Streeter on, 215-16, 218 

Wcseoit and Hon on, 177, 179 
witnesses of, 312, 313 
Alexanian, J. M., 117nl05 
Alford, H., 174, 329 
Allen, T. W„ 18n23 
Allen, W. C, 182n24 
Allibone, S. A.. 162n60 
Alter, F, K., 168 
Ambrose of Milan, 131, 332 
Ambrosiaster, 131, 266, 339 
Amerbach, B., 50 
Ammonius, 322 
Amphoux, C.-B., 73n35, 280nl7, 298n72, 

Amulets, 94n74, 296, 299 
Ancient versions, 52, 94-126. See also 

Anglo-Saxon versions; Arabic versions: 

Armenian version; Coptic versions; 

Ethiopic version; Georgian version; 

Gothic version; Latin versions; Old 

Nubian version; Old Slavonic version; 

Sogdian version; Syriac versions 
Anderson, A., 91 
Andrew of Caesarea, 36, 261 
Andrieu.J., 250nl 
Anglo-Saxon versions, 125-26 
Anli-Mareionite prologues, 33-34n51, 39 
Antiochian text, 161, 279. See also 

Byzantine text 
Aphou, Bishop of Oxyrhynchus, 

Apotropaic magic, 296, 299 
Arabic versions. 122-23, 166 
Arbache, S., 122nll4 
Aristarchus of Samothrace, 198 
Aristophanes of Byzantium, 44, 198 
Aristotle, 44, 197, 199 
Armenian version, 117-18, 165. 166, 311, 319, 

322, 325 
Armstrong, E., 150n30 



General Index 

Artistic adornment, 43-44 

Asceticism, 294-95 

Asiatic manuscript groups, 159, 161 

Athanasius, 67n25, 128, 129, 131, 187, 239n71, 

278, 332 

Augustine, St., 24-25, 95n76, 102, 131, 

202-3, 334 
Aune, D. E., 295n6l, 296n62, 297n67 
Autographs, 180, 281 
Aymon.J., 108 

Baarda, Tj., 119, 134nl34, 134nl37, 152n36, 

Balgue, M„ 57nl2 
Baljon, J. M. S., 229 
Balogh, J., 23n28 
Barbour, R., 3nl 
Bardy, G., 40n60 
Barnard, P, M., 277nl0 
Barnet, D. P., 53nl 
Barns, J. W. B„ 45n66, 57nll 
Basil the Great, 91, 129, 130, 222, 239n71, 

279, 332 
Batiffol, P., 84 
Bauer, W„ 282, 283 
Baumgarten, P, M„ 109n91, 158 
Baumstark, A., 134 

Baur, F. C, 56 

Bayle, P., 206n2 

Beare, F. W., 335n58 

Beatty, C, 53 

Beatty papyri, 41, 42, 53-56, 111, 112, 

133-34, 185 
Becker, U,, 319n33 
Bedier, J., 210-12 
Beermann, G, 83 
Bekker, I., 206-7 
Bell, H. I., 56 
Belser, J., 305n7 
Belsheim, J., 89n60 
Bengel.J. A., 158-61, 165, 169, 

231, 259 
Bensly, R. L., 99 
Bentiey, J., 64nl8, 65n22, 139n5, 142nll, 

Bentiey, R., 156-57, 160, 171, 227 
Benzley, J. D., 12nl0 
Berger, S., 140n9 
Bertsch, L„ 158n51 
Betz, H. D., 296n62 
Bevenot, M., 208n3 
Beza, T., 70, 151-52, 164, 203 
Bieler, L., 205nl, 229n53 
Bilinearity, 18 

Bilingual manuscripts, 45-46 
Bioi, 39-40 
Birch, A., 168-69 
Birdsall.J. N., 57nl2, 83n45, 119, 200n7, 

Birks, T. R., 182n23 
Black, D. A., 231n58 

Black, M„ 172n8, 192n46, 193 194, 331 

Blake, R. P., 119, 310 

Blanchard, A„ 14nl3 

Bludau, A., 146nl9, 266n20, 282 

Blumenkranz, B., 271n28 

Bodmer, Martin, 56 

Bodmer papyri, 41, 56-59, 60, 192n44, 
249n87, 270 

Bogaert, P.-M., 68n26 

Bohairic version, 110. 112, 113, 154, 161, 166, 
168, 179, 187, 319, 327 

Boismard, M.-E., 57nl2, 273nl 

Bolgiani, F„ 132nl31 

Bonner, C, 54n5 

Bonus, A., 182n24 

Booker, C. M„ Un8 

Book-hand writing style, 17-18 

Book of Armagh. See Codex Dublinensis 

Books of Fate, 297 

Boreel, J.. 74 

Bornemann, F. A., 329 

Borner, C. F., 75 

Bornkamm, H„ 145nl8 

Bover, J. M„ 148, 185n29, 190, 329, 333n57. 
336, 338 

Bowyer, W., 162, 164. 229, 230n57 

Brandt, P.-Y., 143nl3 

Branton.J. R., 49n77 

Bray, W. D., 49n77 

Breaden, R. P., 138n2 

Breathing marks, 256 

Brevior lectio potior, 212-14 

Briere, M., 119 

British and Foreign Bible Society, 
191, 336 

Brock, S. P., 40n6O, 126nl21 

Brogan.J., 129nl27, 239n71 

Brooks, J. A., 129nl27, 130. 239n71 

Brown, R., 273nl 

Brown, R. E.. 147n25 

Brown. T„ 153n39 

Browne, G. M„ 124nll6, 124nll7, 297n66 

Bruce, F. F„ 333n57 

Brugensis, F. L., 204 

Bruggen, J. van, 219n28 

Buck, H. M.,Jr..49n77 

Budge, E. A. W., 110n94 

Bunyan.J., 300 

Burgon, J. W„ 181-83, 218, 250nl, 326n45 

Burk, P. D.. 160 

Burkitt, F. C, 102, 175, 183n24 

Burrows, M., 281nl9 

Burton, P., 101n83, 309 

Byzantine recension, 165 

Byzantine text, 52, 212, 215, 218-22, 237-38, 
315, 318, 335. See also Koine text; 
Syrian text 
error or discrepancy in, 265 
INTF project on. 247^18 
rise and development of, 279-80 
witnesses of, 236-37, 280, 306-10 

General Index 


Byzantine Text Project, 249 
Byzantine Text-Type and New Testament 
Textual Criticism, The (Sturz), 221 

Cadbury, H.J., 6lnl5 
Caesarean text, 52, 190, 307, 315, 331, 341 
Streeter on, 215, 218 
witnesses of, 310-12 
Callimachus, 12n9 
Calvin, J., 70, 151 
Cameron, R., 94n74 
Canart, P., 68n26 
Canonical order, 53 
Caracalla, 26n36 
Carey, M., 164 
Carpianus, 39 
Carroll, J. T„ 330n49 
Carson, D. A., 219n27 
Carter, J., 315n31 
Cassiodorus, 29-30, 32 
Catenae, 43 

Cavallo, C, 4nl, 30n43, 84n47 
Celsus, 291-92, 294 
Ceolfrid, Abbot of Jarrow and 

Wearmouth, 106 
Cerny, J., 4n2 

Champlin, R., 74n36, 84n46 
Chapter divisions, 34-36 
Charles I, King of England, 67, 153 
Chaucer, G., 242 
Chaytor, H. J., 23n28 
Childers, J. W., 1 19 
Christian apologia, 290-94 
Christian VII, King of Denmark, 168 
Chrysostom, J., 77, 131, 168. 222, 279, 

296n63, 319 
Cicero. 45, 138n2, 212, 213 
Claremont Profile Method, 236-37, 238-39 
Clark, A. C, 89, 143nl5, 212-14, 305, 309 
Clark, E., 295n59 
Clark, G. H., 219n27 
Clark, K. W„ 143nl6, 260nl2 
Classical method textual criticism, 205-9 
Clement VIII, Pope, 109, 156 
Clementine, 147, 148 
Clement of Alexandria, 91, 128. 131, 166, 179, 

265, 277nl0, 322, 334 
demons, J. T, 100n82 
Codex (introduction of), 12-16 
Codex 1, 148-49 
Codex 16, 92 
Codex 33, 179 
Codex 109, 259 
Codex 157, 89, 90, 310 
Codex 579, 83 
Codex A, 319 
Codex Alexandrinus, 34, 36, 58, 67, 84, 

113n98, 153, 159, 163, 177, 280, 318 
Codex Amiatinus, 106, 707, 108 
Codex Angelicus, 77 
Codex Argenteus, 115-16 

Codex Athous Dionysiou, 86 
Codex Athous Laurae, 84-85 
Codex Augiensis, 74 
Codex Basiliensis, 37, 74 
Codex Beratinus, 84 

Codex Bezae, 74, 80, 99, 103, 125, 128, 150. 
151, 163, 178, 203, 277, 283-84, 288, 292, 
298, 307, 309-10, 317, 318, 320 

description of, 70-73 

error or discrepancy in, 251, 257, 268 

example of, 72 

women depicted in, 290 
Codex Bobiensis, 102 
Codex Boemerianus, 75-76 
Codex Boreelianus, 74 
Codex Borgianus, 80 
Codex C, 319 
Codex Campianus, 77 
Codex (^antabrigiensis. See Codex Bezae 
Codex Cavensis, 106 
Codex Claromontanus, 73-74, 151, 163, 178, 

257, 317 
Codex Coislinianus, 76 
Codex Colbertinus, 103 
Codex Corbiertsis, 103 
Codex Cyprius, 77 
Codex D, 292 
Codex Demidovianus, 168 
Codex Dinaiticus, 6l 
Codex Dublinensis, 81, 106 
Codex F.phraemi, 21-22, 69-70, 113n98, 

172, 179 
Codex Fuldensis, 108 
Codex Gigas, 94, 103-4 
Codex Harleianus, 74-75. 108 
Codex Koridethi, 83, 215, 310 
Codex Laudianus, 74, 75 
Codex Mediolanensis, 108 
Codex Monacensis, 81 
Codex Montfortianus, 147 
Codex Mosquensis, 77, 78, 80 
Codex Mutinensis, 76 
Codex Nitriensis, 79-80 
Codex Ottobonianus, 147 
Codex Palatinus, 102 
Codex Petropolitanus, 83-84 
Codex Porphyrianus, 79 

Codex Purpureus Petropolitanus, 77-79 

Codex Ravianus, 147n25 

Codex Regius, 77, 147, 179 

Codex Rossanensis, 44, 84, 85 

Codex S, 269 

Codex Sangailensis, 82-83, 108 

Codex Sangermanensis, 74 

Codex Sinaiticus. 8, 19, 55, 68, 81, 113, 172, 

173, 179, 182, 234-35n64, 278, 318 
chapter divisions in, 34 

estimated cost of producing. 26 
example of, 66 
history of, 62-67 
production of, 15-16 


General Index 

Codex Sinopensis, 44, 79 

Codex Tischendorfianus III, 83 

Codex Vaticanus, 8, 59, 77, 80, 86, 113, 154, 

167, 182, 184, 278, 318, 322 
chapter divisions in, 36 
error or discrepancy in, 252, 253, 254, 260, 

264, 265, 271 
history of, 67-69 
production of, 15-16 
Codex Vaticanus 2066, 86 
Codex Vaticanus (B), 34, 83, 169, 179, 312 
Codex Vercellensis, 102 
Codex Veronensis, 102 
Codex Washingtonianus, U3n98 
Codex Wolfii A. See Codex Harleianus 
Codex Wolfii B, 76 
Codex Zacynthius, 43, 83, 173 
Coimfort, P. W., 231n59 
Cola, 45-46 

COLLATE program, 242, 243 
Collins, A., 155n42 
Colophons, 29, 31-33. See also Jerusalem 

Colwell, E. C, 43n62, 47n73, 49n77, 57nl2, 

191n42, 210, 233-35, 236, 238, 239, 

279nl5, 306nll 
Comfort, P. W„ 53nl 

Comma Johanneum, 146-48, 162, 182n23, 219 
Commata, 45-46 
Commentaries, 41-43 
Complutensian Polyglot Bible, 139-42, 146, 

148, 149, 150 
Comprehensive Profile Method, 238-39 
Computers in textual criticism, 240-46 
Conflation of readings, 176, 180, 265, 279-80 
Conjectural emendation, 226-31 
Conrad, E. W., 94n74, 296n64 
Constantine, Emperor, 15-16, 68 
Constantinopolitan group, 161-62, 166, 169 
Contamination of text, 208-9. See also Error 
Contractions, 17, 23-24 
Conybeare, F. C, 97, 281n20 
Coptic versions, 110-15, 179 
Cornelius Tacitus, 50-51 
Correctors, 25 
Cosgrove, C. H., 330n49 
Cosmas of Prague, 103 
Council of Trullo, 22 
Cozza-Luzi, G., 68 
Cramer, J., 229 
Creed, J. M„ 89n62, 326n43 
Crispin, J., 151 
Critical Examination of the late New 

Testament A, (Twells), 158 
Cronin, H. S., 89n60 

Cross, F. L, 36n53, 143nl3, 143nl6, 185n29 
Croy, N. C, 326n44 
Crum, W. E., 127nl23, 297n69 
Cureton, W., 96, 98 
Curetonian Syriac version, 96, 97, 178, 268, 

277, 287, 309 

Cursive writing style, 17-18, 87, 252-53 
Cyprian of Carthage, St., 102, 126, 131, 

170, 178 
Cyril, 128, 179, 187 
Cyril, St., 121-22 
Cyrillic alphabet, 121 
Cyril of Alexandria, 131, 166, 332, 334 
Cyril of Jerusalem, 129, 327 

Da Bassano, F, 121 

Dain, A., 3nl, 223n39 

Daly, M., 289n42, 289n44 

Damascus, Pope, 101, 105, 201 

Damste, A., 229n53 

Daniels, B. L, 236n67 

Danove, P. L., 322n39 

Davidson, R„ 172n8 

Davies, D. M„ 120nl08 

Davies, D. W., 152n35 

Davies, N. de Garis, 5n4 

Dearing, V. A., 241n75 

De Bruyne, D., 33n50, 105 

Defence of the New Testament (Bengel), 160 

De Hamel, C, 8n5, 9, 31n45, 44n62 

Deissmann, A., 56 

De Koe, S. S., 229 

Delitzsch, F., 140n9 

Delobel, J., 276n7, 309 

Demosthenes, 45 

Denaux, A., 274n3 

Denzinger, H., 148n26 

Descent of Manuscripts (Clark), 213 

Devil's Bible. See Codex Gigas 

De Vine, C. F, 332n54, 333n57 

Devreesse, R., 87n53 

Dewey, A. J., 94n74 

Diatessaron, 108, 117, 131-34, 268, 287, 295, 

302, 310, 319 
Dibbelt, H„ I45nl8 
Didymus the Blind, 128, 129, 130, 131, 179, 

187, 278, 320n34, 332 
Diglot New Testament, 191 
Diiler, A., 18n23 
Diocletian, Emperor, 25-26, 81, 

123, 220 
Dionysius, 44, 179 
Dissemination of literature, 274-76 
Dittography, 254 
Doctrinal disputes, 282-87 
Donaldson, A. M., 292n53 
Dopp, S„ 40n60, 130nl30 
Doran, M., 228n52 
Dorotheus of Tyre, 40 
Drexl, F., 298n71 
Ducas, D., 139n7 
Dudik, B., 103n87 
Dummer, J., 297n66 
Dupont, D. J„ 340n64 

Eadfrid, Bishop of Lindisfarne, 125 
Eastern recension, 161-62 

General Index 


Ecclesiastical text, 279. See also Byzantine text 
Eclecticism, 222-26 
Edwards, E. G., 316n32 
Egypto-Palestinian manuscript group, 161 
Ehrman, B. D., 18n23, 22n25, 47n73, 53nl, 
101n83, 110n93, 117nl05, 119, 126nl21, 
128nl24, 129nl27, 130, 214n22, 218n26, 
222n36, 225n46, 231n60, 235n64, 
238n70, 239n71, 240n74, 251nl, 260nl2, 
264nl8, 265nl9, 277n8, 280nl7, 281nl9, 
281n20, 282n23, 283n26, 284n31, 
285n32, 285n33, 286n34, 286n35, 
286n36, 292n53, 302-3n2, 308nl6, 
312n28, 320n34, 333n55, 333n57 
Eichhorn, J. G„ 87 
Eliot, J„ I63n64 
Ellens, J. H., 197nl 
Elliott, J. K„ 64nl8, 100n83, 114nl00, 222n36, 

223n38, 224, 225n46, 235n65, 26lnl6 
Elliott, W. J., 248 
Ellis, A. A., 156n46 
Ellison, J., 240 

Elzevir texts, 152, 154, 155, 160, 164 
Emendation, 205, 226-31 
Emmanuel (manuscript copyist), 88 
Enslin, M. S., 91n69 
Ephesus, 56 
Ephraem, St., 21, 69, 97-98, 131, 133-34, 

183n24, 268 
Ephraim, 91 
Epicurus, 44 

Epiphanius, St., 129, 131, 239n71, 295n60, 332 
Epp, E. J., 53nl, 231n58, 231n59, 236n67, 251nl, 

260nl2, 274n4, 281n20, 283-84, 288 
Erasmus, D., 88, 142-50, 203 
Erotemata (Lascaris), 138n2 
Error, 184, 250-71 

addition of details, 268-71 

addition of natural complements, 263-64 

agreements in, 232 

from faulty eyesight, 251-54 

from faulty hearing, 254-57 

from historical or geographical 

difficulties, 264 
intentional changes, 259-71 
of judgment, 258-59 
of the mind, 257-58 
primitive, 229 

of spelling and grammar, 261-62 
unintentional changes, 251-59 
Ersch, J. S„ 206n2 
Estienne. See Stephanus 
Ethiopic version, 119-21, 161, 166, 322 
Euclid, 199 

Eumenes II, King of Pergamum, 8 
Euripides, 44 

European Old Latin manuscripts, 101, 102-5 
Eusebian canons, 38-39, 69 
Eusebius of Caesarea, 15-16, 38, 39, 67n24, 
83, 91, 131, 166, 173, 187, 199, 215, 220, 
265, 311, 322, 327, 328 

Eusebius of Vercelli, 102 
Euthalian apparatus, 39-40 
Euthalius, 40, 76, 79 

External evidence, 302, 305-13, 318, 320, 325, 
328, 329, 332, 334, 335, 338, 339, 342 

Falla, T. C, 96n78 

Family 1, 87, 91, 215, 310, 311, 341 

Family 13, 87-90, 106, 215, 264, 310, 311, 341 

Family 1424, 90-92 

Family relationships, determining, 231-39 

Farmer, W. R., 323-25n41 

Farrer, A. M., 326n43 

Fairer, J., 206n2 

Farstad, A. L„ 219n28, 220, 221, 249 

Fascher, E„ 260nl2, 281n20 

Fayyumic version, 110, 112 

Fee, G. D„ 59nl4, 66n23, 126nl21, 129, 

231n58, 231n59, 239n71, 251nl, 264nl8, 

289-90, 312n28 
Felix V (anti-pope), 143nl3 
Fell, J., 153-54 
Ferrar, W. H., 87 
Ferrar group, 87, 106, 233 
Fiorenza, E. S„ 289n43 
Fischer, B., 105, 109, 240n73, 241n75 
Fischer, J., 241 
Fitzmyer, J. A., 288n39 
Flacius, M., 206 
Flack, E. .E„ 39n59 
Flavius Josephus, 103 
Fleury palimpsest, 102 
Flock, G., 41n61 
F6rster, H., 115 
Fortune-telling, 295-99 
Four Gospels, The (Streeter), 214-15 
Fox, A., 154n4l, 157n47 
Frank, T, 26n35 
Franklin, B., 164 
Franssen, H., 229 

Frederick Augustus, King of Saxony, 62 
Freer, C. L., 80, 202 
Frey, J., 81n44 
Friend, A. M., Jr., 27n40, 44 
Froben, J., 50, 142 
Froude, J. A., 146 
Froy (friar), 146 
Fuller, D. O., 219n27 
Funk, R. O., 132nl33, 256n8 
Furlan, I., Iln8 
Fust, J., 138n2 

Gager, J., 287n37 

Gaine, H., 164 

Galen, 14, 199 

Gallagher, E., 291n50, 291n51, 292n52 

Gamble, H. Y„ 5n3, 8n5, 13nl2, 14nl3, 

24n30, 274n6 
Gardner, A., 30n44 
Gardthausen, V., 3nl, 31n45 
Garitte. G., 119 


General Index 

Garrett, S. R., 296n62 

Garzaniti, M., 121nlll 

Gaumer, T., 192n45 

Gaventa, B., 330n49 

Geanakoplos, D. J., 139n7 

Geerlings, J., 74n36, 84n46, 87n54, 259nll 

Geerlings, W., 40n60, 130nl30 

Genealogical method, 176, 181, 210-12, 

235n66, 302 
Geneva Bible, 151n34, 203 
Georgian version, 118-19, 311, 319, 320, 

322, 327 
Gerstinger, H., 54n4 
Gibson, M. D., 100 
Gilbert of Porree, 203 
Gilliam, J. F., 132nl33 
Gilliard, F. D., 23n28 
Gilman, I., 125nll9 
Giorgi, A. A., 80n41 
Glagolitic alphabet, 121 
Glaue, P., 115nl04, 253n5 
Glazier, W. S., 114 
Globe, A., 284n31, 310, 3Hn22 
Glosses, 41, 125, 258 
Goguel, M., 140n9 
Golden Gospels, 108-9 
Gonis, N., 6lnl6 
Goodacre, M., 134nl37, 242n78 
Goodspeed, E. J., 43n62, 230n57 
Gore, C, 223n41 
Gorzaga da Fonseca, L., 253n5 
Gospels of Mac-Regol. See Rushworth Gospels 
Gothic version, 115-16, 154, 165, 266, 319, 327 
Gow, J., 228n51 
Granfield, P., 6lnl5, 270n25 
Grape, A., 116 
Graser, E. R., 26n35 
Gray, H. L., 88n56 
Gray, L. H., 125nll8 
Greber.J., 71n32 
Greek-English Diglotfor the Use of Translators, 

A (Kilpatrick), 224-25 
Greek New Testament 

first printing of, 137-39 

manuscripts of, 52-94 

in the modern critical period, 165-94 

in the precritical period, 137-64 

scientific textual criticism of, 165-70 

statistics of, 47-5 1 

variant readings collection, 152-64 
Greek New Testament According to the 
Majority Text, The (Hodges & Farstad, 
eds.), 220 
Greenlee, J. H., 83n45, 185n29 
Greeven, H., 119, 336n59 
Gregory, C. R„ 17n20, 48, 49, 81n44, 83, 
109n90, 143nl5, 147, 150n32, 156n44, 
172-73, 184n28, 186, 189 
Gregory, Pope, 106 
Gregory of Nazianzus, 131 
Gregory of Nyssa, 90, 129, 130, 131, 318 

Grenfell, B. P., 55 

Gressmann, H., 6lnl5 

Gribomont, J., 84n47 

Grierson, R., 98n81 

Griesbach.J. J., 165-68, 173, 175, 188, 303 

Gruber, J. G., 206n2 

Gruber, L. F., 90-91 

Griinberg, M„ 126nl20 

Gryson, R,, 101n85, 109, 115nl03 

Guidi, I., 122 

Gutenberg, J., 137-38 

Gwilliam, G. H„ 99, 182n24 

Gwynn.J., 99 

Hadrian, 56n9 

Haines-Eitzen, K., 24n30, 25n33 

Hall, f. W.. 205nl 

Hall, I. H., 164, 194n48 

Hannah, D. D., 129nl27, 239n71 

Hannick, C, 121nlll 

Haplography, 254 

Harclean version, 96, 99-100. 166. 277, 309 

Harding, S., 106 

Harley, R., 75, 108 

Harlfinger, D., 3nl 

Harmonization, 184, 262-63, 303, 313-14, 335 

Harms, R„ 49n77 

Harrington, D., 282n23 

Harris, J. R.. 4n2, 26. 59nl5. 90n65, 125nll8, 

146n23. 253n6, 268n23, 270n26, 281n20, 

283, 287 
Harris, W. V., 14nl4 
Harrison, B., 181n22 
Hailing, D., 229 
Hartman, I. I., 229n53 
Harwood, E., 163 
Haseloff. A.. 84n47 
Haskins, C. H„ 88n56 
Hatch, W. H. P., 17n22, 28n40, 34n52, 36n53, 

40n60, 43n62, 55n7, 76n37, 86n51, 

96n78, 149n28, 307nl4 
Havet, L., 205nl 

Head, P. M., 27n38, 260nl2, 284n31, 303n2 
Headlam, A. C, 182n24 
Hebbelynck, A., Illn94 
Heinsius, D., 152n36 
Hellegourac'hJ., 50n79 
Helm, K„ 115nl04 
Helps for readers, 33-47, 149 
Hemmerdinger, B., 18n23 
Henderson, C, Jr., 93n73 
Hendricks, 1. H. M., 5n3 
Hendrickson, G. L., 23n28 
Hennecke, E., 6lnl5 
Ilenrichs, A., 94n74, 296n64 
Henry, F„ 93n73 
Hcracleon, 277, 308 
Hermas, 23n29 
Herodotus, 5 
Hesychian recension, 187 
Hesychius of Egypt, 169, 187 

General Index 


Hcizenauer, M., 305n7 

Hexapla (Origen), 200 

I lieronymus, Sophronius F.usebius. 
See Jerome, Si. 

Hilary of Poitiers, 131, 170, 334 

Hills, E. F, 182n23, 219 

Hippolytus of Rome, 131, 178, 286 

Hjelt, A., 98 

Hodge, C, I67nl 

Hodges, Z. C, 219n28, 220. 221, 249 

Hoeg, C, 46n71 

Hoclzlin, J., 152n36 

Holmes, M. W„ 18n23, 22n25, 47n73, 53nl, 
101n83, 110n93, 117nl05, 119, 126nl21, 
129nl27, 214n22, 218n26, 221n34, 
222r>36, 225n46, 239n71, 240n74, 251nl, 
260nl2, 264nl8, 280nl7, 303n2 

Homer, 79, 110, 112, 197-98 

Homoeotelcuton, 253, 254, 303 

Home, T. H., 174 

Horner, G., 32n46, 114 

Hon, F.J. A., 171, 174, 175-84. 187, 188, 190, 
191, 193, 212, 213, 215, 229, 244, 279, 
280, 282, 302, 303n3, 306, 307, 308, 309, 
312, 315, 329, 332n54, 333n57, 336, 338, 
339n6l, 340n64, 342 

Hoskier, H. C, 89, 109n90, 188n34, 
269, 270 

Housman, A. E„ 226, 305n8, 315 

Howe, C. J., 235n66, 238n69, 245n80 

Hug, J. L., 142nl0, 169, 215 

Hulbert-Powell, C. I.., I60n54 

Hulley, K. K„ 201nl4 

Hultgren, A. J., 282n22 

Hunger, H., 3nl, 57nl0 

Hunt, R. W., 143nl3 

Hunzinger, C.-H., 32846 

Hurtado, L. W.. 80n42, 235n64, 311 

Husselman, E. M., 112, 114 

Hutton, E. A., 233 

Hyatt, J. P., 57nl2 

Hypertext, 245—46 

Hypotheses, 39-40, 149 

Hyvernal, II., Ill, 114 

IGNTP. See International Greek New 

Testament Project 
1-H-Kie-a, 187 

Iliad (Homer), 79, 197-98, 213 
[mmerwahr, H. R., 12nl0 
ImnaisVili, I., 119 
Ink making, 10-11 
Instantius, Bishop of Spain, 148 
Institut fur neutestamentliche Tcxtforschung 

(INTF), 241, 242, 246, 247-48 
Intergroup profile, 239 

Internal evidence, 302-4, 313-15, 318-19, 325, 
328, 332, 334, 341, 342 

of documents, 176 

of groups, 176 

of readings, 175-76 

International Greek New Testament Project 

(IGNTP), 241, 242, 244, 246, 248-49 
INTF. See Institut fur neutestamentliche 

Intra-group profile, 239 
Intrinsic probability, 175—76, 184, 

303-4, 341 
Introduction to the Critical Study and 

Knowledge of the Holy Scriptures 

(Home), 174 
Irenaeus, 33, 91, 131, 170, 178, 265, 277, 286, 

308, 327 
lrigoin, J., 211nl3 
Isaac, J., 287 
Isaac the Great, 117 
Isidore of Pelusium, 131, 166 
Isidore of Seville, 103 
Itacism, 255 

James, M. R., 88n56 

James I, King of England, 74 

Jammers, E., 46n72 

Janeras, S., 277n8, 308nl6 

Jebb, R. C, 205nl, 227n50 

Jenson, N., 140n8 

Jeremias,J., 331, 342n67 

Jerome, St., 11, 45, 81, 101, 102, 103, 105-6, 
121, 127nl22, 129nl26, 131, 137, 138, 
145, 151, 156, 169, 170, 201-2, 203, 220, 
260nl3, 266, 267n21, 304, 322, 327, 332 

Jerusalem colophon, 83, 89, 90, 310 

Jerusalem recension, 187 

Jewish-Christian relations, 287-88 

Job, M., 119 

John II Comnenus, Emperor, 88 

Johnson, E. E., 330n49 

Jones, I,. W., I6nl8 

Jonge, H.J. de, I46n22, 152n36 

Joose, P., 134nl37 

Juckel, A., 99 

Judge, E. A., 94n74, 296n64, 296n65 

Jiilichcr, A., 105 

Julius II, Pope, 142nl0 

Junack, K., 47n73, 193, 250nl 

Junginann, J. A., 6lnl5, 270n25 

Junius, F., 151 

Justin Martyr, St., 128, 131, 178, 277, 286, 
308, 323 

Kaczynski, R., 296n63 
Kadzaia, L, 119 
Kahle, P. F.„ 112n95, 323n40 
Kammerer, W., 114 
Kannaday, W., 291 
Kantorowicz, H., 205nl 
Karavidopoulos, J., 194 
Karnetzki, M., 260nl2 
Kasser, R., 58, 112, 114 
Katz, P., 13nll 
Kay, W. A., 190n40 
Kenney, E. J., 209 


General Index 

Kenyon, F. G., 53, 55n6. 56, 67, 110-lln94, 

143nl5, 213, 340n65, 341n65 
Ker, N. R., 126nl20 
Kessler, H. L., 44n63 
Khalatheants, G., 117 
KibiskiJ., 44n63 
Kilpatrick, G. D., 45n66, 190n38, 191, 222n38, 

224-25, 231n58, 26lnl6 
King James Version, 145, 146, 152, 157, 181. 
218-19, 317, 323, 327, 334, 338 

error or discrepancy in, 254, 255, 257-58, 
262-63, 264, 271 

subscriptions in, 41 
Kiraz, G. A.. 96n78 
Kitto, J., 150n32 
Kleist, J. A., 45n68 
Klijn, A. F. J., 73n34, 95n77, 152n36 
Klimkeit, H.-J., 125nll9 
Knox, B. M. W., 23n28 
Koenen, L., 94n74 
Koenigs, F., 185 

Koine text, 52, 187, 188, 212, 279, 315, 335, 
336, 338. See also Byzantine text 

forms antedating, 307-10, 318 

witnesses of, 306-10 
Koriun, Bishop, 117 
Koster, B., 191 
Kraeling, C. H., 132nl33 
Kraft, B., 187n33 

Kraft, R., 240n74, 241, 282n21, 287n37 
Kramer, B., 24n30 
KransJ., 230n56 
Kraus, T. J., 54n4 
Krchnak, A, 143nl3 
Kremer, J., 308nl7 
Krodel, G., 282n21 
Kriiger, F, 187n33 
Kruger, M. J., 94n74 
Kuster, L, 154 

Lachmann, K., 163, 170-71, 173, 174, 175, 

206-7, 329, 339, 342 
Lactantius, 138n2 
Lagrange, M.-J., Illn94, 203nl9, 222n37, 

307nl3, 310n21 
Lake, K.. 3nl, 15nl6, 17nl9, 64, 85, 86, 87n54, 

90, 184n28, 188n34, 279, 281n20, 283, 

310, 333n57 
Lake, S., 3nl, 17nl9, 84n46, 87n54 
Lamouille, A., 273nl 
Lampe, G. W. H., 24n31 
Landgraf, A., 2()3nl9 
Lanfranc, 106 
Lascaris, C, 138n2 
Latin versions, 100-109 
Latin Vulgate, 39, 102, 103, 117, 125, 127, 138, 

139, 140, 142, 145, 146, 150, 151, 153, 

156, 167, 168, 189, 203, 269, 304-5, 327 
apperance of Comma Johanneum 

in, 148 
description of, 105-9 

first printing of, 137 

Lachmann's use of, 170 
Lauch, E., 64nl8 
Laud, Archbishop, 74 
Laurence, R., I67nl 
Lazar of Pharb, 117 
Leach, B., 4n2 
Leal, J., 266n20 
Leaney, R., 90n64 
Leclerc.J., 160,214 

Lectionaries, 20, 35, 4(>-Al, 48-49, 50, 258 
Lectors, 25, 26, 46 
Legg, S. C. E., 191-92 
Lehmann, H., 117 

Leloir, D. L., 117nl05, 133nl34, 134 
Leo X, Pope, 139, 140, 142nl0 
Leo XIII, Pope, 148 
Letis, T. P.. 219n27 
Letter of Aristeas, 341 
Lewis, A. S., 96, 98, 100 
Lewis, C. T„ 159n53 
Lewis, J., 242 
Lewis, N., 5n3, 14nl4 
Liber apologeticus, 147^18 
Lieberman, S., 13nll 
Lietzmann, H., 188n34, 304n6 
Ligatures, 138 

Lightfoot, R. H., 326n43, 329 
Lindisfarne Gospels, 108, 125 
Lindsay, W. M., 33n49 
Liuzza, R. M., 126nl20 
Local genealogical method, 235n66 
Local texts, 169, 203, 214-18 
Loerke, W. C, 28n40, 84n47 
Lohmeyer, E., 326n43 
Lombard, P., 203 

Longenecker, R„ 59nl4, 235n64, 312n28 
Longo, O., Iln8 

Lord's Prayer, 90, 113-14, 162, 167, 262 
Lor'k'ip'anidze, K., 119 
Lowe, E. A., 1 ln8, 102, 108n89, 109n90 
Lucar, C, 67, 153 
Lucas, A., 4n2 

Lucian of Antioch, 161, 169, 187, 215 
Lucifer of Calaris, 104, 131, 170, 332 
Lunellum, 10 
Luther, M, 145, I63n64 
Luthi, K. J.,8n5 
Lyell, J. P. R., 139n6 
Lyon, R. W., 69n28 
Lyonnet, S., 189 

Maas, P.. 205nl, 208, 211nll 
Mabillon, }., 206 
Mace, D., 157-58, 330 
Macgregor, G. H. C, 333n57 
Macomber, W. F, 120 
Madan, E, 29n42 
Magdeburg Centuriators, 206 
Magic, 295-99 
Majority text, 218-22, 279 

General Index 


Majuscules, 17-21, 41, 43, 46, 48, 177 

error in, 231 

example of, 19 

Gricsbach's classification of, 166 

important, 52, 62-86 

Lachmann's use of, 170 

number of imponant, 50 

Wettstein's apparatus in, 161 
Malherbe, A.. 330n49 
Malzer. G„ 158n51 
Mani Codex. 94 
Mann, T., 301 
Manson, T. W., 191n42 
Manuscript 28, 87 
Manuscript 33. 87-88 
Manuscript 61, 88 
Manuscript 69, 88 
Manuscript 81, 88 
Manuscript 157, 88-89 
Manuscript 0169, 92-93 
Manuscript 383. 89 
Manuscripr 461, 92 
Manuscript 565, 89 
Manuscript 579, 89 
Manuscript 614, 89 
Manuscript 700, 89-90 
Manuscript 892, 90 
Manuscript 1071. 90 
Manuscript 1241, 90 
Manuscript 1424, 91 
Manuscript 1582. 91 
Manuscript 1739, 91 
Manuscript 2053. 91 
Manuscript 2344. 92 
MANUSCRIPT program, 243 
Manutius, A., 140 
Marazuela, T. A., 148n27 
Marcan prioriry, 293, 304 
Marchand, J. W., 40n60 
Marcion, 90. 131, 178. 187, 189, 199n3, 266. 

277. 282-83n24, 305, 308 
Marcionile prologues. 39 
Marcus, J„ 292n54, 293n58 
Marsden. R., 126nl20 
Marshall, T., 154 
Martin, J. P. P., 327n45 
Martin, V., 56, 57. 58 
Martini, C. M., 59nl4, 68n26, 189. 192n46, 

19X 308, 312n28 
Matthaei. C. K, 167-68 
Matthew, H. C. G., 181n22 
Matzkow, W„ 105 
Mazon. P., 223 

McArthur, H. K„ 36n53, 38n57 
McCartney, E. .S., 23n28 
McGing, B. C., 54n4 
McGurk, P., 33n51, 93n73 
McKendrick. S., 51n80, 190n40 
McLachlan. H , 158n50 
McNamee. K ( . 17 
McReynolds, P.. 236 

Medici, Cardinal de, 142nl0 

Medici, Catherine de. 92 

Mees, M., 278nl0 

Menander, 44 

Menologian, 47 

Menoud, P.-H., 58nl3, 284n29, 288n40 

Merk, A., 148, 185n29, 189-90, 329, 333n57, 
336, 338 

Merx, A.. 304, 309 

Mesrop, St., 117 

Methodius, St., 121-22, 128 

Metzger, B. M., 28n40, 28n41, 39n59, 45n69, 
49n77, 50n78, 51n80, 6lnl5, 90n64, 
92n71, 94n75, 114nl00, 114nl01, 
115nl03, 120, 121nlll, 122nll2, 
123nll5, 127nl22, 129nl25, 129nl26, 
139n5, 142nll, 147n25, 151n34, 162n61. 
I64n67, 190n39, 192n45, 192n46, 193, 
200n7, 201nl3, 2O3n20, 214n21, 230n57, 
258n9, 260nl4, 270n25, 298n72, 306n9, 
311n24, 340n65, 341n65, 342n66 

Meyer, A., 6lnl5 

Meyer, K„ 32n48 

Michael (monk), 80 

Michaelis, W., 26lnl6 

Michclsen, J. 11. A., 229 

Middle Egyptian version, 110, 113, 114 

Middleton, C, 156n45 

MigneJ. P., 30n44, 201nl5, 260nl3 

Mill, J., 154-55, 157-58, 159, 164 

Miller. K., 182n24, 250nl 

Miller. J. E„ 259nl0, 290n45 

Miller, P. N\, 153n37 

Milne, II. J. M., 27n39, 65 

Milton, J., 227 

Miner, D., Iln8, 109n9O 

Minervius, 201 

Mink, G.. 110n93, 112, 235n66 

Minuscules, 17-21, 43. 48, 148-49, 
169-70, 177 
error in, 251 
example of, 20 

Gricsbach's classification of. 166 
important, 52, 86-92 
number of important, 50 
Wettstein's apparatus in, 161 

Moffatt, J„ 230n57, 329 

Moldenhauer, D. G., 168 

Molitor, J„ 98, 119, 337n6() 

Monk, J. H., 156n45, 227n50 

Monophysite controversy, 283 

Montanus, A., 153 

Montfaucon, B. de, 206 

Morrell, M. K., 126nl20 

Morrill. B., 242 

Morton, A. Q., 65n20 

Moses of Chorion, 117 

Moss, C, 100 

Most, G. W., 198n2 

Moulton.J. H., 333n56 

Muehler, II., 4nl 


General Index 

Mullen, R., 129nl27, 130, 239n71, 249 
Multiple readings, 233 
Munoz, A,, 84n47 
Musurillo, H., 208n3 
Mynors, R. A. B., 30n43 

Nanni, G., 206 

Nau, F., 252 

Nazianzen, G., 138nl 

Necropolis of Thebes, 5 

Nestle, Eberhard, 96, 143nJ5, 173, 185n29, 

190-91, 229n53, 230, 336, 341n65 
Nestle, Erwin, 190, 191 
Nestle-Aland text, 191, 193, 230n56, 231, 244, 

247, 248 
Neumes, 46 
Neutral text, 177, 179-80, 184, 187, 213, 

215, 307 
Neville, Archbishop of York, 88 
Neville, C, 175nll 
Nevius, R. C, 24n30 
New, S., 86, 310 

New English Bible (NEB), 192, 338 
Newing, E, G., 94n74, 296n64 
New Revised Standard Version, 317, 329, 338 
New Testament in Greek and English, The, 157 
New Testament in the Original Greek, The 

(Westcott & Hon), 175 
New Testament in the Original Greek According 

to the Byzantine/Majority Textiform, The 

(Robinson & Pierpont), 219 
Newton, B. W., 174 
Nickelsburg, G. W. E., 287n37 
Nicklas, T., 71n31 
Nicolaus (monk). 18n23 
Nikolopoulos, P., 65 
Nolli, G., 148, 189n37 
Nomina sacra, 23-24 
Nonnus, 319 
Nordberg, H., 67n25 

Nordenfalk, C, 38n55, 38n57, 43n62, 134nl36 
North, J. L, 126nl21 

Oblak, V., 122 
O'Connell, M., 130nl30 
Odyssey (Homer), 197-98 
Oecumenius, 91 
O'Flanerty, W. D., 302n2 
Old Latin versions, 159, 170, 178, 189, 277, 
307, 309, 319, 320, 322, 327, 328, 341 

description of, 101-5 

error or discrepancy in, 266, 269, 270 
Old Mercian dialect, 125 
Old Nubian version, 123-24 
Old Slavonic version, 121-22 
Old Syriac version, 96-98, 277, 309, 31 1 
Oliver, H. H., 39n58, 95n77 
Omanson, R. L, 194 
Omont, H., 79 
Onomastica, 43 
Opisthographs, 12 

Opitz, H. G„ 190n38 

Opiza manuscript, 1 18 

Orelli.J. K., 50 

Origen, 59nl4, 83, 91, 127nl22, 128, 129, 131, 
161, 165-66, 169, 170, 179, 187, 200-201, 
215, 220, 278, 292, 294, 310, 311, 
312n28, 319, 322 

Osburn, C, 47n73, 129nl27, 239n71 

Ostraca, 49 

O'Sullivan, O., 51n80, 190n40 

Ott, W., 241 

Outtier, B., 298n72 

Owen, H., 162n62 

OwenJ., 153n38, 154 

Oxford Vulgate, 109 

Oxyrhynchus papyrus 4499, 61 

Paap, A. H. R. E., 24n30 

Pagels, E., 289n43 

Paleography, 3, 206 

Palestinian Syriac version, 96, 100, 169, 267, 

277, 309 
Palimpsests, 21-22 
Palmer, E., 184-85 
Pamphilus, 67, 76, 187 
Pannartz, A., 138n2 
Panten, K. E„ 73n33 
Papadopoulos-Kerameus, A., 44n64 
Papebroch, D., 206 
Paper, early use of, 10 
Papyri, 4-7, 8, 11-12,48 

important, 52, 53-61 

number of important MS on, 50 

parchment superiority to, 14—15 

writing on, 16 
Parablepsis, 253. 303 
Parassoglou, G. M., 27n39 
Parchment and vellum, 4, 8-10, 14-17, 20, 21, 

48. See also Purple parchment manuscripts 
Parker, D. C, 22n25, 46n70, 61, 70, 73n35. 
83n45, 134nl37, 211nl2, 236n67, 
241n77, 245, 248, 249, 274n3, 280nl7, 
281, 298n72, 310, 322n39 
Parkinson, R., 4n2 
Parsons, E. A., 197nl 
Parsons, M., 284n31 
Parsons, P. J„ 4nl 
Parvis, M. M., 192n43, 233 
Patrick, G. A., 175nll 
Patristic quotations, 52, 126-34, 166 
Paul VI, Pope, 68n26 
Payne, P. B., 289n45 
Pelagius, 131, 332, 334, 339 
Perkins, A., 132nl33 
Perrot, C., 100n82 

Peshitta version, 96, 98-99, 166, 182n24 
Petersen, W., 134nl37, 274n3, 308nl7 
Petrus Ethyops, 121 
Petzer, J., 101n83, 304n4 
Pfeiffer, R., 278nll 
Philo Judaeus, 198 

General Index 


Philoxenian version, 96, 99-100. 165, 169 

Philoxenus, Bishop of Mabbug, 99 

Pickering, S. R., 50n78, 94n74 

Pickering, W., 219n28, 296n64, 296n65 

Piercy, W. C, 223n41 

Pierpont, W., 219, 221, 249 

Pigulevskaya, N. V,, 100 

Pilgrim 's Progress (Bunyan), 300 

Pisano, S., 68n26 

Pius X, Pope, 109 

Plato, 44 

Piatt, T. P., 121 

Pliny the Elder, 5, 7, 8, 12 

Plooi|, D„ 102n86 

Plunkett, M. A., 284n31 

Plutarch, 197 

Polycarp, 99 

Polyglot Bibles, 138, 153 

Port, A„ 89 

Porter, C. L„ 59nl4, 312n28 

Postgate, J. P., 205nl 

Pratorius, R, 121 

Preisendanz, K., 296n64 

Preuschen, E., 201nll, 201nl2 

Primasius, Bishop of Hadrumentum, 131 

Primitive error, 229 

Priscillian, 105, 148 

Prolegomena, 154 

Pseudo-Hieronymus, 131,339 

Ptolemy Epiphanes, 8 

Punctuation, 22, 41, 193 

Purple parchment manuscripts, 11, 44, 77-79, 

84, 89, 102, 108-9, 115-16 
Pusey, P. E., 99 
Puyemre, 5 

Quadrilinearity, 18 

Quantitative method, 234-35, 236, 237, 239 

Quarternion, 30 

QuastenJ., 130nl30 

Quecke, H„ 113 

Quinisext Council, 22 

Quires, 17 

Quirke, S., 4n2 

Rabbula, Bishop of Edessa, 98 

Racine, J.-F, 129nl27, 130, 239n71, 279nl4 

Radermacher, L., 226n48 

Rahner, K., 148n26 

Rand, E. K., I6nl8 

Rapp, C, 30n43 

Rational criticism, 222 

Reasoned eclecticism, 223 

Receasions, 161-62, 165-66, 205 

Redus, M. W„ 49n77 

Reed, R., 8n5 

Regul.J., 33-34n51, 39n59 

Reicke, B„ 143nl3 

Reitzenstein, R„ 127nl23 

Renehan, R., 209 

Rettig, H. C. M., 83 

Reuchlin, 145 

Reuss, E., 149n28, 157n48. 163, 194 

Revised Standard Version, 181 , 329, 338 

Reynolds, L. D., 21 

Rhenanus, IS., 50, 142 

Rhodes, E. F., 57nll, 117, 231n59, 238n68 

Richards, W. I.., 235n64, 236n67 

Rico, M. R., 139n6 

Riddle, D., 43n62, 47n73, 49n77, 281n20, 283 

Riesenfeld, H„ 321n37 

Robens, C. H., 13nll, 13nl2, 14nl3, 24, 

55-56, 93n73, 288n41, 296n64, 341n65 
Robenson, A. T., 150n31 
Robiason.J. A., 40n60 
Robinson, M.. 219, 221, 249 
Robinson, P., 242 
Romer, C, 94n74 

Ropes, J. H., 15nl6, 97, 279, 307nl4 
Rorem, P., 281nl9 
Rosen, E., 251n2 
Rosenbaum, E., 44n63 
Royse, J„ 251nl, 302n2 
Ruerher, R., 287 
Rufinus of Aquileia, 131, 337 
Rules of St. Pachomius, 127nl23 
Ruling patterns, 16-17 
Rupp, E, G., 175nll 
Rushworth Gospels, 108, 125 
Ryder, M. 1... 8n5 
Rypins, S., 79n39 

Sabas (monk), 90 

Sabio, N., 149 

Saenger, P., 23n28, 139n4 

Sahak, 117 

Sahidic version, 1 10-13. 187, 254, 264, 267, 

270, 319, 327 
Sailors, T. I!., 292n53 
Salmon, C, 183 
Sanday, W„ 182n24, 213 
Sanders, H. A., 55n6, 77, 80, 253n5 
Sandys, J. E., 205nl 
Sauer, C, I63n64 
Saunders, E. W., 260nl2 
Scanlin, H. P., 221n34 
SchafersJ., 325n42 
Schaff, P., 194n48 
Schapiro, M., 134nl36 
Schenke, H.-M., 113, 114nl01, 115 
Schenkeveld, D. M., 23n28 
Schmid, J., 79, 91, 224, 252, 26lnl5 
Schmid, W., 226n48 
Schmidt, C, 253n5 
Schmidt, K„ 124 
Schmidtke, A., 89n6l 
Schmiedel, P. W., 229-30 
Schmitz, F.-J., 112, 115 
Schoeffer, P., 138n2 
Scholarly criticism, 197-204 
Scholderer, V„ 138n2, 140n8 
Scholia. 41-43 


General Index 

Scholz.J. M. A., 169-70 

Scholz, P. O., 123nm 

Schone, H., 199n6 

Schubart, W., 56 

Schulthess, I'r., 100n82 

Schulz.J. C. F„ 162n62 

Schiirmann, H., 331 

Scientific textual criticism, 165-70 

Scribes, 16-31 

Scriptio continua, 22-23 

Scriptoria, 25, 30 

Scrivener, F. H. A., 70, 74, 76n38, 88, 143, 171, 

183, 298n70 
Scrolls, 11-14 
Seidel, A. F.., 75 
Sellin, P. R., 152n35 
Semler, J. S., 161-62, 165 
Serapion, 311n22 
Sessa, M., 149 
Sevcenko, I., 64nl8 
Shakespeare, W., 228 
Shapur I, King, 132 
Sharp, J. L, III, 126nl20 
Simon, M., 287 

Simon, R., 204 

Sinaitic Syriac version, 96-97, 175, 267, 277, 

295, 304, 309, 322, 341 
Sixtine Vulgate, 109 
Sixtus V, Pope, 109 
Skeat. T. C. 13nl2, 14, 15nl6, 24, 25n34, 53. 

54n4, 65, 68, 297n66 
Smith, J. P., 189 
Smith, M„ 296n62 

Social history of Christianity, 280-99 
Sogdian version, 124-25 
Soncino press. 1 37 
Sophocles, 44 
Sophronius, 40 
Sorles Astrampsychi. l^l 
Souter, A., 184^85, 188n34, 188n35, 

I91n42, 213 
Sozomen, 261 
Sparks, H. F. D„ 109 
Speculum, 104-5 

Spencer, M., 235n66, 238n69, 245n80 
Sporotno, M. V., 139n5 
Spyridon, Bishop, 261 
Statistical analysis, computerized, 245 
SteinmillerJ. E., 109 
Stemma, 207-8, 216 
Stendahl, K„ 328n46 
Stephanus, H., 151 
Stephanus, R., 149-52, 153, 154. 

184, 185 
Stesimbrotus of Thasos, 197 
Sticbos, 25-26 
StojkovicJ., 143nl3 
Stone, M. E.. 1 18 
Stonehouse, N. B., 326n43 
Sireane, A. W.. 174 
Strecker. G.. 282n21, 282n23 

Streeter, B. H„ 79, 84, 88, 90, 91, 169, 203. 

214-18, 278nl2, 302. 310 
Streitberg, W., 116 
Strugnell, J., 231n58 
Stunica, D. L. de, 139, 146 
Sturz, H.. 221-22. 279nl6, 306n9 
Stutz, E., 115nl03 

Sub-Achmimic version, 110, 112. 113. 319 
Subscriptions, 40-41 
Suggs, M.J. , 236n67 
Superscriptions, 40-41 
Swanson, R. J., 259nlO, 278nl() 
Sweynheym, C, 138n2 
Swift, J. , 155n42 
Synaxaria, 47 
Synopsis. 38 

Syriac versions, 96-100, 151. 161, 189. 327 
Syriac Vulgate. See Peshitta version 
Syrian text, 177-78, 180-81, 183. 187, 215. 

279. 306. See also Byzantine text 

Talismans, 49-50 
Tarelli, C. C. 143nl6 
Tasker, R. V. G., 192 

Tatian, 97, 108, 117, 131-34, 178. 187. 188. 
189, 190, 26lnl7, 266. 268, 287, 295. 
302, 305, 308, 310, 319, 323, 327 
Taylor, D. G. K., 50n78, 94n74, 134nl37 
Taylor, M, 287n37 
Tbet' manuscript, 118-19 
Tcnney, M., 59nl4, 235n64, 312n28 
Tertullian, 131, 178, 265, 277, 308 
Teststelhn, 237-38, 247 
Testuz, M„ 58 
Teubner, B. G., 208 
Text division, 53 
Textual Commentary on the Creek New 

Testament, A, 193 
Textus Receptus, 89, 127, 145, 153, 159, 162, 
170, 215. 219. 221, 232-34, 249, 323. 327. 
329, 342 

abandonment of. 155, 156-57. 167 

first L : .S. publication of. 163-64 

history of, 149-52 

overthrow of, 170-94 
Theagenes of Rcgium, 197 
Theile, W„ 167, 304n5 
Theodore of Mopsuestia, 131. 328 
Theodoret. 132 
Theodore the Studite, 30 
Thcodotians, 199-200 
Thcodotus, 199 
Theodulf, 106 
Tbeotogicat Tendency of Codex Bezae 

Cantabrigiensis in Acts, 7fr<?(Epp), 283 
Theophrastus, 199 
Thomas, I, Jr., 163-64 
Thomas of Harkel, 99 
Thompson, E. M.. 3nl, 67, 253n5 
Thompson, H.. Illn94. 112, 114 
Thomson, L. S , 32n47 

General Index 


Thomson, R. W., 200n7 

Thoroughgoing eclecticism, 222-26 

Thorpe, J. C, 245n80 

Tillyard, H. J. W., 46n71 

Tindall, C, 65n20 

Tischendorf, L. F. C. von, 62-64, 69, 74, 77, 

79, 80, 147n25, 156n44, 172-74, 190, 230, 

329, 336, 338, 340n63 
Tisserant, E., 92n72 
Titles, 36, 48 
Titus, E. L., 259nl2 
Tkacz, C. B., 106n88 
Trajan, 56n9 
Transcriptional probability. 175-76, 184, 

302-3, 341 
Traube, L., 24n30 

Tregelles, S. P., 83, 173-74, 230, 329 
Tremellius, E., 151 
Triphyllios of Ledra, 261 
Triple readings, 233 
Tselos, D. T., 44n63 
Tune, E. W., 234n63 
Turner, C. H., 40n60, 146n23, 223-24 
Turner, E. C, 14nl3 
Turner, E. G., 4nl, 4n2 
Turyn, A., 3nl 
Twells, L., 158 
Tychsen, O. G., 168 

Uhlig, S., 120nl08, 120nll0 

Ulfilas, 115, 116 

Uncials, 17 

United Bible Societies, 192-94, 247, 249, 336 

United Bible Societies' Greek New Testament, 

191, 193-94 
Ussher, Archbishop, 153 

Vaganay, L., 210 

Vajs.J., 122 

Valentine-Richards, A. V., 89n62 

Van den Ven, P., 229n53 

Van der Sande Bakhuyzen, W. H., 229 

Van der Valk, M, 327n45 

Van Friesen, O., 116 

Van Kampen, K., 126nl20, 139n4 

Van Manen, W. C, 229 

Van Rompay, L., 96n78 

Van Unnik, W. C, 33n50, 152n36 

Variant readings 

collection of, 152-64 

criteria for evaluation of, 300-305 

process of evaluating, 305-15 
Velleius Paterculus, 50 
Vellum. See Parchment and vellum 
Verginius Rufus, 12 
Vergotte, J., 4n2 

Vernacular languages, Bible in, 137, 138 
Vetus Latina project, 241 
Victor, Bishop, 108 
Victor I, Pope, 199 
Victor of Capua, 132nl31 

Vikgren, A., 95n77 

Vinaver, E., 250nl 

Vincent, M. R., 142nl0, 159n53 

Vogel, M., 31n45 

Vogels, H. J., 148, 185n29, 189, 287, 298n70, 

329, 333n57, 336, 338 
Vogt, E., 328n46 
Volumes, production of, 11 
Von der Goltz, E., 91 
Von Dobschiitz, E., 40n60, 143nl5, 329 
Von Gebhardt, O., 84, 168 
Von Harnack, A., 281n20, 283, 304-5, 333n57 
Von Ostermann, G. F., 138n3 
Von Schlozer, A. L., 209 
Von Soden, H., 36n53, 40n60, 79-80, 86, 90, 

91, 134, 143nl5, 185-89, 236, 279, 

297n68, 305, 329, 336, 338 
Voobus, A., 94n75, 98n80, 192n46 
Voskresenski, G. N'., 122 
Voss, D. C, 320n36 

Wachtel, K., 18n23, 235n65, 235n66, 238n69, 
241n77, 245n80, 279nl5, 306nl0 

Walker, W. O., 273n2 

Wallace, D. B., 218n26, 221n34, 331n33 

Walton, B., 121, 153, 154 

Walzer, R., 199n4 

Warren, M., 27n38 

Wartenberg, U., 4n2 

Washington codex of Pauline Epistles, 77 

Wattenbach, W., 32n48, 109n90 

Weber, R., 109 

WeimaJ., 330n49 

Weinrich, W. C, 45n69 

Weiss, B., 183-84, 190, 223, 230, 329, 333n57, 
336, 338, 339 

Weitenberg, J. J. S., 117 

Weitzmann, K., 44n63, 44n65 

Wellesz, E. C}., 46n71 

Wellhausen, J., 331 

Wells, C. B„ 132nl33 

Wells, K., 155 

Weitzmann, K., 43n62 

Wendel, C, 15nl6 

West, M. L., 198n2, 208-9 

Westcott, B. F., 171, 174-84, 187, 190, 191. 
193, 213, 215, 229, 244, 279, 280nl8, 
282n24, 302, 303n3, 308, 309, 329, 
333n57, 336, 338, 339n6l, 340n64, 342 

Western non-interpolations, 179-80, 309 

Western order, 53, 70, 80, 115 

Western recension, 162, 165-66 

Western text, 52, 190, 213, 221, 237-38, 284, 
315, 327, 329, 331, 337, 338, 339, 341 
description of, 307-10 
rise and development of, 276-77 
Wescott and Hon on on, 177, 178-79 
witnesses of, 277, 310-12, 318 

Wettstein, J. J., 48, 160-61, 162 

WhareyJ. B., 301nl 

Whibley, L., 205nl 


General Index 

Whiston, W., 71n32 

Whitby, D., 154-55 

White, H. J., 108, 109 

White, J., 99 

Whittingham, W., 151, 203 

Wifstrand, A., 27n39 

Wikgren, A., 192n46, 193, 194 

Wilcken, U„ 56 

Wilkenhauser, A., 252 

Wilkins, D., 168 

Williams, C. R., 326n45 

Williams, C. S. C, 260nl2, 281n20, 333n57 

Williams, J. F., 326n43 

Willis, W. H., 93n73 

Willoughby, H. R, 43n62 

Wilson, D. W. G„ 21 

Wilson, F. P., 228n52 

Wilson, J. M., 71n32 

Wilson, N. G., 3nl 

Winer, G. B., 172 

Winslow, M. W., 86 

Wire, A. C, 289n43, 290n46 

Wisse, F., 236, 237 

Witherington, B., 290 

Witnesses, 52-134, 302. See also Ancient 
versions; Greek New Testament, 
manuscripts of; Patristic quotations 

Wolf, F., 206 

Wolf, J. C, 75, 76 

Women, oppression of, 288-90 

Wordsworth, J., 108, 109 

Wright, W., 150n32 

Wyman, L. C, 86 

Ximenes de Cisneros, F, 139-42, 146 

Yoder, J. D., 73n34 

Zeno, 44 

Zenodotus of Ephesus, 198 
Zereteli. G., 18n23 
Zetzel.J. E. G„ 129nl26 
Zigabenus, F.., 319-20 
Zimmer, F., 329 
Zohrabian, Y., 117 

Zunu, G., 25n33, 40n60, 212nl6, 221n32, 224, 
227n49, 274n5, 279nl6, 308nl7, 312n27 
Zuurmond, R., 119nl07, 120nl09 

Index of Biblical Passages 

Page numbers set in italics refer to figures. 

Matt. 1.2 


Malt. 26.3 

Matt. 1.8 


Matt. 27.3-5 

Matt. 1.16 


Matt. 27.9 

Matt. 1.17 


Matt. 27.16-17 

Matt. 1.21 


Matt. 27.41 

Matt. 2.6 


Matt. 27.49 

Matt. 3.15 


Matt. 28.7 

Matt. HI 


Mark 1.1 

Matt. 5.18-24 


Mark 1.1-5.30 

Matt, 6.4 


Mark 1.1-10.22 

Matt. 6.6 


Mark 1.2 

Matt. 6.9-13 


Mark 1.10 

Matt. 6.13 


Mark 1.15 

Matt. 6.20-8.5 


Mark 1.41 

Matt. 6.27 


Mark 2.9-14 

Matt. 9.13 


Mark 2.26 

Matt. 11.16 


Mark 3.5 

Matt. 12.36 


Mark 4.28 

Matt. 12.47 


Mark 5.3 

Matt. 13.5-10 


Mark 5.11 

Matt. 13.14-16 


Mark 5.31-16.20 

Matt. 13-24 


Mark 5.41 

Matt. 14.22-31 


Mark 543 

Mart. 15.8 


Mark 6.3 

Matt. 15.38-16.7 


Mark 6.34 

Matt. 16.2-3 


Mark 8.2 

Matt. 17.26 


Mark 8.31 

Matt. 18.1 


Mark 9.17 

Matt. 18.11 


Mark 9.29 

Matt. 19.17 


Mark 10.17 

Matt. 20.28 

71, 84, 268 

Mark 10.19 

Man. 20.30 


Mark 10.40 

Man. 21.19 


Mark 10.41 

Matt. 21.23 


Mark 11.14 

Matt. 21.28 


Mark 11.27 

Malt. 22.1 


Mark 13.11 

Matt. 22.4 


Mark 13.32 

Man. 22.7 


Mark 14.6 

Man. 22.34-35 


Mark 14.25 

Mati. 23.14 


Mark 14.31 

Matt. 24.36 


Mark 14.62 

Matt. 2531 


Mark 14.65 



202, 264 






80, 310 












80, 311 








268, 295 
















Index of Biblical Passages 

Mark 15.25 
Mark 15.34 
Mark 16.3 
Mark 16.6-20 
Mark 16.8 
Mark 16.9-20 
Mark 16.14 
Luke 1.1-8.12 
Luke 1.1-11.33 
Luke 1.3 
Luke 1.15-28 
Luke 1.35 
Luke 2.14 
Luke 2.33 
Luke 2.36 
Luke 2.41 
Luke 2.43 
Luke 2.48 
Luke 2.51-3.7 
Luke 3.22 
Luke 3.23-38 
Luke 5.5 
Luke 5.26 
Luke 5.32 
Luke 5.38-69 
Luke 6.4 
Luke 6.20-30 
Luke 7.31 
Luke 8.13-24.53 
Luke 9.51 
Luke 10.1 
Luke 10.25 
Luke 10.32 
Luke 11.2 
Luke 11.2-4 
Luke 11.3-4 
Luke 11.12 
Luke 11.13 
Luke 11.32 
Luke 11.35 
Luke 12.9 
Luke 12.25 
Luke 14.17 
Luke 14.26 
Luke 14.27 
Luke 16.9-21 
Luke 16.19 
Luke 16.25 
Luke 17.35 
Luke 18.18 
Luke 20.1 
Luke 21.14 
Luke 21.32 
Luke 21.38 
Luke 21.45-47 
Luke 21.50-53 
Luke 22.15-20 
Luke 22.19-20 
Luke 22-23 
Luke 22.43-44 
Luke 22.44-56 





113, 325-26 

120, 322-27 


80, 313 






267-68, 285 




267-68, 285 


128, 284n32, 285 









80, 306 


40, 340-42 
















253n5, 270 







87, 212 




180n20, 285n32 


284n31, 284n32, 286 


Luke 22.61-63 
Luke 23.32 
Luke 23.34 
Luke 23.38 
Luke 24.3 
Luke 24.4-5 
Luke 24.18 
Luke 24.31-33 
John 1.1-4.40 
John 1.1-5.11 
John 1.1-6.11 
John 1.1-8.38 
John 1.1-839 
John 1.1-10 
John 1.1-14 
John 1.13 
John 1.18 
John 1.28 
John 1.34 
John 1.35-38 
John 2.1 
John 2.17 
John 2.19 
John 4.1 
John 4.22 
John 5.3b-4 
John 5.7 
John 5.7-8 
John 5.12-21.25 
John 539 
John 6.4 
John 6-8 
John 6.30-41 
John 6.35b-14.15 
John 6.50-8.52 
John 7.8 
John 7.37-39 
John 7.52 
John 7.52-8.12 
John 7.53-8.11 
John 8.9 
John 8.12 
John 8.39 
John 10.7 
John 14-21 
John 17.15 
John 19.10-16 
John 1914 
John 19.17-35 
John 1920 
John 19.35 
John 1936 
John 20.31 
John 21 
John 21.1-25 
John 21.24 
John 21.25 
Acts 1.14 
Acts 4.25 
Acts 6.8 
Acts 7.46 


267, 294 










310, 313, 314n30 





284n32. 286 

201, 264 






















87, 162, 319-21 






















Index of Biblical Passages 


Acts 8.30 
Acts 8.36-38 
Acts 8.37 
Acts 9.5-6 
Acts 9.6 
Acts 10.30 
Acts 12.10 
Acts 12.25 
Acts 13.32 
Acts 13.43 
Acts 15.20 
Acts 15.28 
Acts 15.29 
Acts 15.34 
Acts 15.40 
Acts 16.12 
Arts 16.27 
Acts 17.4 
Acts 17.12 
Acts 19.9 
Acts 19-34 
Acts 19.40 
Acts 20.28 
Acts 25.13 
Acts 26.14-15 
Acts 28.1-13 
Rom. 1-3 
Rom. 1.27-2.15 
Rom. 1.32 
Rom. 4.5-5.3 
Rom. 4.12 
Rom. 5.1 
Rom. 5.6 
Rom. 5.8-13 
Rom. 6.5 
Rom. 7.2 
Rom. 7.14 
Rom. 8.1 
Rom. 8.32 
Rom. 12.11 
Rom. 13.3 
Rom. 13-4 
Rom. 13.9 
Rom. 14.17 
Rom. 15.29-33 
Rom. 15.32 
Rom. 16.21 
Rom. 16.25-27 
1 Cor. 3.9 
1 Cor. 7.5 
1 Cor. 9.2 
1 Cor. 11.2-10 
1 Cor. 12.2 
] Cor. 12.13 
1 Cor. 14.34-35 
1 Cor. 15.51 

1 Cor. 15.54 

2 Cor. 3.3 
2 Cor. 317 

2 Cor. 4.13-12.6 
2 Cor. 6.4 



74, 162 





229n54, 339-40 





73, 203 







73, 203 



229n54, 265, 331-33 
































259nl0, 289 







2 Cor. 6.14-7.1 
2 Cor. 7.8 
2 Cor. 8.4 
2 Cor. 12.7 
Gal. 2.12 
Gal. 4.18 
Gal. 4.19 
Gal. 4.25 
Gal. 4.28 
Gal. 5.1 
Gal. 6.17 
Eph. 1.7 
Eph. 34 
Eph. 4.14 
Phil. 2,1 
Phil. 4.3 
Col. 1.2 
Col. 1.12 
Col. 1.14 
Col. 1.23 
Col. 2.2 
Col. 2.18 
Col. 2.23 
Col. 4.16 
1 Thess. 2.7 

1 Thess. 3.2 

2 Thess. 1.10 
2 Thess. 2.14 
1 Tim. 2.1-10 
1 Tim. 3.16 

1 Tim. 4.3 
1 Tim. 5.3-16 

1 Tim. 6.7 

2 Tim. 1.13 
2 Tim. 2.24 
2 Tim. 4.19 
Phitem. 9 
Heb. 1.3 
Heb. 1.8 
Heb. 2.9 
Heb. 4.2 
Heb. 4.11 
Heb. 533 
Heb. 7.1 
Heb. 9.4 
Heb. 9.14 
Heb. 10.1 
Heb. 11.4 
Heb. 11.37 
Heb. 12.11 
Heb. 12.20 
Heb. 13-21 
Jam. 2.18 

1 Pet. 1.1-3 
1 Pet. 1.3 
1 Pet. 17 
1 Pet. 1.12 
1 Pet. 2.11 
1 Pet. 2.21 
1 Pet. 37 
1 Pet 318 





















229n54, 334-35 




256, 328-30 
























229n54, 230 














Index of Biblical Passages 

1 Pet. 3.19 


Rev. 2.12 

1 Pet. 3.21 

229n54, 256 

Rev. 2.13 

1 Pet. 5.10 


Kev. 2.20 

2 Pet. 2.13 


Rev. 31 

2 Pet. 2.18 


Rev. 3.7 

2 Pet. 310 


Rev. 3.14 

2 Pet. 312 


Rev. 3.19-4.1 

1 John 2.23 


Rev. 4.3 

1 John 4.3 


Rev. 4.8 

1 John 5.7-8 

88, 162, 182n23, 219 

Rev. 9.2-3 

1 John 5.10 


Rev. 9.10 

2 John v. 8 


Rev. 11.3 

2 John v. 12 


Rev. 13.10 

Jude 1 


Rev. 13.15 

Jude 5 


Rev. !3.l6 

Jude 22 


Rev. 15.6 

Kev. 1.4 


Rev. 18.12 

Rev. !.5 

255, 262 

Rev. 18.14 

Rev. 1.6 


Rev. 19.13 

Kev. 1.15 


Rev. 22.18 

Rev. 1.20 


Rev. 22.18-19